IEEE/CSC & ESAS European Superconductivity News Forum (ESNF), No.

6, October 2008
(ASC Preprint 2AP01 conforming to IEEE Policy on Electronic Dissemination, Section 8.1.9)
The published version of this manuscript appeared in IEEE Transactions on Applied Superconductivity
19, No. 3, Part 2, 1055 - 1068 (2009)

Next Generation More-Electric Aircraft:

A Potential Application for HTS
Superconductors
Cesar A. Luongo, Senior Member, IEEE, Philippe J. Masson, Senior Member, IEEE, Taewoo Nam,
Dimitri Mavris, Hyun D. Kim, Gerald V. Brown, Mark Waters, David Hall

stringent weight and volume constraints imposed by an

Abstract—Sustainability in the aviation industry calls for airborne application. Through our multi-disciplinary
aircraft that are significantly quieter and more fuel efficient than collaboration we were able to assess, for the first time, the
today’s fleet. Achieving this will require revolutionary new feasibility of non-conventional aircraft architectures based on
concepts, in particular, electric propulsion. Superconducting
superconducting electric propulsion. In the next sections we
machines offer the only viable path to achieve the power densities
needed in airborne applications. This paper outlines the main provide case studies used to assess the prospects of
issues involved in using superconductors for aeropropulsion. We superconductivity on aircraft propulsion and show that
review the work done under a 5-year program to investigate the superconducting machines have already achieved power
feasibility of superconducting electric propulsion, and to densities comparable to turbine engines. To fully enable
integrate, for the first time, the multiple disciplines and areas of electric flight however, power densities need to improve even
expertise needed to design electric aircraft. It is shown that
further, which is only possible with all-superconducting
superconductivity is clearly the enabling technology for the more
efficient turbo-electric aircraft of the future. machines. We developed design concepts for revolutionary
aircraft using superconducting machines for propulsion and
Index Terms— Aircraft, electric propulsion, superconducting showed that with further development in superconducting and
motor cryocooling technologies, all within reach, superconductivity-
enabled flight could be a reality within the next 20 years.

I. INTRODUCTION II. MOTIVATION

R ELENTLESS growth in air traffic poses the question of

how to achieve a more sustainable way of keeping
The major drivers in the design of future aircraft are: 1)
reduced airport noise, 2) reduced emissions (both pollutants
humankind flying without further environmental damage, and and greenhouse gasses), and 3) reduced fuel burn. NASA has
especially in a fuel-efficient manner. This paper discusses the set performance goals for these corners of design trade space
merits and challenges associated with the revolutionary as enumerated in Table I.
conversion of current day aeropropulsion to novel schemes
based on turbo-electric propulsion. Superconducting rotating TABLE I
machines (motors and generators) are destined to play a key NASA SUBSONIC AIRLINER PERFORMANCE GOALS*
role in this conversion as the enabling technology that will
Corner of the N+1 (~ 2015) N+2 (~2020) N+3 (~2030)
allow this conversion to electric propulsion within the very trade space Conventional Unconventiona Advanced Aircraft
Tube & l Hybrid Wing Concepts
Manuscript received August 26, 2008. Wing Body (relative to user defined
This research supported by the NASA Fundamental Aeronautics Program (relative to (relative to reference)
and the Department of Defense Research and Engineering (DDR&E) division B737/CFM56) B777/GE90)
under the URETI on Aeropropulsion and Power. Support also received from Noise 55 LDN at average
Florida State University’s Center for Advanced Power Systems (CAPS) -32 dB -42 dB
(below Stage 4) airport boundary
C. A. Luongo is with the Dept. Mechanical Engineering, Florida A&M-
Florida State University College of Engineering, Tallahassee, FL, USA (e- LTO NOx Better than
mail: [email protected]). Currently on sabbatical with ITER IO, CEA Emission (ref. -60 % -75 % -75%
Cadarache, St. Paul-lez-Durance, France CAEP 6)
P. J. Masson is with the Center for Advanced Power Systems, Tallahassee,
FL, USA. Currently with the Advanced Magnet Lab, Palm Bay, FL, USA (e- Performance: Better than
mail: [email protected]) Aircraft Fuel -33 % -40 % -70%
T. Nam and D. Mavris are with the Aerospace Systems Design Laboratory, Burn
Dept. of Aeronautical Engineering, Georgia Institute of Technology, Atlanta, Performance: Exploit Metroplex
GA, USA (e-mail: [email protected]) -33 % -50 %
Field Length Concept (STOL)
H. Kim and G. Brown are with NASA Glenn Research Center, Cleveland
OH, USA (e-mail: [email protected], [email protected]) * National Aeronautics and Space Administration Headquarters, NASA
M. Waters and D. Hall are with DHC Engineering, San Luis Obispo, CA, Research Announcement (NRA): NNH08ZEA001N, Mar. 7, 2008
USA (e-mail: [email protected])
ESNF, No. 6, October 2008; ASC Preprint 2AP01 conforming to IEEE Policy on Electronic Dissemination, Section 8.1.9

need to achieve sustainable aviation in the face of projected air

traffic growth, conservatively predicted to double over the
next 20 years (see Fig. 1).
Over the last 50 years the aircraft industry has responded to
increased air traffic through a process of technology adoption
and improvement that has consistently reduced noise levels
and fuel consumption. Fig. 2 shows the decrease in fleet
average noise as a result of improvements driven by tighter
noise regulations (referred to as “stages”). This tightening of
regulations is expected to continue as airports become
engulfed by suburban sprawl. Likewise, overall fuel efficiency
has steadily improved as evidenced in Fig. 3 showing the fuel
consumption per pound of thrust for jet engines since their
introduction in the 1950s. Note that fuel efficiency has nearly
Fig. 1. Air traffic growth (in trillions of passengers-km). (from E. L. Gervais, doubled through the introduction and improvement of high
"Boeing Commercial Airplanes Current Product Overview," presented at San bypass turbo-engines. Note the successive “step changes” in
Diego County Regional Airport Authority Advisory Committee Meeting, San efficiency with the introduction of a new engine type.
Diego, CA, Jul. 12, 2007)
A fair question to ask is whether the required performance
23 Concorde goals outlined in Table I can be achieved with the
conventional high bypass ratio turbofan engine. The answer to
? this question may be that new basic propulsion concepts are
10.0 needed and much of the current research sponsored by NASA
Stage 2
Average Noise is addressing this prospect. Some “new” ideas are not new at
Level Relative B-737-200
To Stage 3
DC9-10
B-727-200
Negotiated
Out of Service
all, such as the use of compressor intercooling and
Ave
(EPNdB) In S rage Stage 3
erv
recuperation of exhaust heat to pre-heat air into the
0.0 B-747-100 ice Stage 4
B-727-100 B-747-200
MD-800
B-747-300
A300-600R
combustors. These are “old” ideas, but they are worth re-
A300
A310-222
757-200
A330-300
A 10 dB cumulative
evaluating in light of heat exchanger materials and designs.
B 777-200
747-400

MD-90-30
reduction from Stage 3 However, one technology that is definitely new, at least for
-10.0 use in aircraft engines, is a propulsion system design that uses
1960 1970 1980 1990 2000 2010 2020 advanced superconducting, cryogenically cooled electric
Year of Certification
Fig. 2. Aircraft fleet average noise levels continue to decrease. (from F. generators and motors to drive a multitude of low noise
Collier, E. Zavala, and D. Huff, "Fundamental Aeronautics Program Subsonic electric fans. The obvious break-through that must be achieved
Fixed Wing Project Reference Document," NASA) for this to happen is a marked increase in the power to weight
ratio of electric generators and motors.
A brief discussion on the advantages of electric
aeropropulsion is presented next. Then to demonstrate the
potential of an electric propulsion system based on HTS
technology, we present a number of examples for propulsion
systems as well as for full aircraft of N+3 generation (first in
service circa 2030).

III. ADVANTAGES OF ELECTRIC AEROPROPULSION

A. Present-day high bypass turbofans
The majority of today’s large civil transports are propelled
by high bypass ratio engines. This type of engine deploys a
large fan mechanically connected and driven by low pressure
turbines as illustrated in Fig. 4. Most power produced by the
core engine is used to rotate the fan, which in turn will
generate most of engine thrust. Turbofans can be very
Fig. 3. Specific fuel burn rate (lb/hr of fuel consumption per lb of thrust) has
been steadily decreasing since the introduction of jet engines in the 1950s. compact with specific power in the range of 3-8 kW/kg. The
(from H.D. Kim, J.J. Berton, and S.M. Jones, "Low Noise Cruise Efficient bypass ratio (BPR), defined as the ratio of the mass flow rate
Short Take-Off and Landing Transport Vehicle Study," AIAA-2006-7738, of the stream passing outside the core divided by that of the
Sept. 2006.) stream flowing through the core, plays a key design parameter
Future generations of aircraft are designated as N+1, N+2, of the engine. A higher BPR, in general, yields lower exhaust
etc. (with N being today’s technology). Note that some of speed, which serves to reduce fuel consumption and engine
these targets are particularly aggressive, but are driven by the noise at the cost of an increase in weight and fan diameter. In
ESNF, No. 6, October 2008; ASC Preprint 2AP01 conforming to IEEE Policy on Electronic Dissemination, Section 8.1.9

addition, any increases in an engine’s BPR must be i.e., the replacement of traditionally pneumatically and
accompanied by matching improvements in core specific hydraulically driven functions with electrically powered
output power so as to maintain fan pressure ratio (the ratio of components [2]. Overall, the electric propulsion scheme opens
air pressure in front of the fan to that after it) within a up the aircraft design space to many new possibilities in which
reasonable range. Prodigious advancement in specific power major leaps can be made towards achieving the performance
of engine core and material for the past several decades goals specified in Table I.
resulted in a significant increase in BPR. Recent engines such
C. Electrical Ducted Fan Concept
as the GE90 turbofan exhibit a BPR of 9:1. Nevertheless, the
practical upper limit to increasing the BPR of modern turbofan In order to take advantage of electrical propulsion, electrical
engines is fettered by the inherent coupling between propulsor “propulsors” have to be designed. We can take advantage of
operation mechanics and core thermodynamic cycle [1]. the high bypass ratio of current engine and replace the engine
core with an electrical motor. Thrust would then be generated
To Airplane
Hydraulic
System
To Airplane Bleed Air
System (Environmental
through the fan rotation only as shown in Fig. 6.
Control System, APU,
etc.)

Strut

To Fan Air-Cooled
Gearbox- Bleed air
mounted cooler
To ATS Fan produces
Hydraulic
Fan Pump
Pressure-
>85% of thrust
Regulating Valves
Exhaust
Nozzle
Jet produces
<15% of thrust

To ATS Combustor High

High Pressure Pressure
Low Compressor Turbine
Pressure Low
Gearbox:
Compressor
Generator(s)
Pressure
Turbine

Hydraulic Pump

Lube Pumps

Air Turbine Starter (ATS)
Cowl

Fig. 4. High bypass ratio turbofan

B. The Case for Electric Propulsion a. Conventional turbine and fan coupled on the same shaft

In a traditional aero turbine engine, the fan and turbine

engine are mounted on the same shaft and rotate at the same
speed. The traditional arrangement is shown in Fig. 5a.
Turbines naturally exhibit higher efficiency, and are more
power-dense, at higher speed; their application to aircraft
propulsion greatly limits their rotation speed as the tip of the
fan’s blades should not go supersonic except for maximum
power operations such as take-off when fuel efficiency is not a
major driver. This normally means that aero turbine rotation
speeds are limited to a few thousand RPM. Moreover, torque
and speed are coupled in turbofans, limiting any potential
efficiency gain through speed control.
Fig. 5.b illustrates a notional example of how HTS motor
technology can help relax this coupling. Each generator is b. Turbine and fan decoupling through electrical converters
mechanically linked to a corresponding turboshaft, whereas
the propulsor (could be a propeller or a fan) is electrically Fig. 5. Comparison of conventional and turbo-electric aeropropulsion
connected to the generators through an “electrical gearbox”
Large Aircraft Turbine Engine Large Aircraft Electric Drive
Decoupling torque and speed would lead to very valuable High Bypass Fan Section with electrical motor

control flexibility to enable a more favorable trade between

on-design and off-design performance. In addition, this Bypass Fan
Section of Aircraf t
Engine
architecture is intrinsically compatible with the emerging
concept of “distributed propulsion” that produces thrust by Electrical
Superconducting
Drive Motor

means of multiple small propulsors or engines embedded on

the wing or fuselage. This revolutionary airframe/propulsion Superconducting
Motor Replaces
Turbine
integration concept can be achieved with the proposed
High Bypass Turbofan Electrical Ducted Fan
architecture by remotely connecting multiple propulsors to a .

generator. This arrangement is anticipated to surpass other Fig. 6. Electrically driven propulsion system: electrical ducted fan
distributed propulsion concepts in many aspects, as will be
further discussed in section VII. Furthermore, this type of Such a system is feasible only if electrical motors can be of
hybrid architecture is intrinsically well-suited with the about the same size or better than aero turbines. Conventional
emerging and continuing trend towards more electric aircraft; motors exhibit a specific power up to 0.5 kW/kg, too low
ESNF, No. 6, October 2008; ASC Preprint 2AP01 conforming to IEEE Policy on Electronic Dissemination, Section 8.1.9

compared to turbine engine cores. Therefore conventional technology for turbo-electric aeropropulsion. Unfortunately
machines, which are limited by heat generation in copper fully superconducting machines using HTS conductors are not
windings, cannot be used for aircraft propulsion. A new yet practical, until low AC loss HTS conductors are fully
technology has to be considered to lift that specific power developed.
limitation. Superconducting machines offer the only hope of
E. Cooling Considerations
ever achieving electrical aeropropulsion.
Since superconductors require operation at cryogenic
D. High Temperature Superconductivity as Enabling temperature, the cooling system is a very important part of the
Technology to Electric Aeropropulsion aircraft architecture. The configuration and feasibility of the
Superconductors can carry very high current density with superconducting machine depends on the cooling system and
no resistance thus enabling very light machines. High specific mostly on the cooling power available.
power has been demonstrated in superconducting machines
1) Cryocoolers
since the late 60s. A few low temperature superconducting
machines have been built with both the stator and rotor Cryocoolers are an obvious choice as they represent an
superconducting thanks to the availability of NbTi low AC active source of cooling using a close loop of cryogen.
loss conductors. Cryocoolers are “plug and play” reliable systems and therefore
appropriate for airborne application. However, by today
100
Industrial motors Industrial motors standards, they are still too heavy for use on aircraft.
PEM Fuel Cells @ 2 kW/kg
Turbofan engines, w/o propulsors
Cryo generator tested, AF/Westing.
outweigh turbine engines
Cryo motor 5 M$ design, AF/Westing.
AF/Westinghouse cryo Al fit
by factor of 5 to 10. Off-the-shelf cryocoolers exhibit efficiencies of about 10-
10 Cryo designs (more elec. rocket Long)
HTS rotor & stator (approx.), AF/OSU 15% of Carnot efficiency, which correspond to about 70W/W
HTS gen. designs @ 10 krpm, Masson
All Bi2223 Masson, wi Fe, EM + 60% at 30 K. The lightest cryocoolers today weigh about 5 lb/HP-
W eight (klb)

1
Turbine engine core input (or 3 kg/kW-input) as shown in Fig. 8 (based partially on
data from Ref [3]). This is just for the cold head portion, the
associated compressors and ancillaries represent an overhead
0.1 All-Superconducting of about 5 times that weight. The use of packaged turbo-
motors may be lighter
by a factor of 3. compressors may reduce this overhead significantly, and
0.01
coupled with the development of much lighter cold heads, it
0.1 1 10 100 may be possible to reach the target of 3 kg/kW-input as overall
Shaft Power and Equivalent Shaft Power (kHP) specific weight for cryocoolers by the time N+3 aircraft come
into service (~2030-2035).
Fig. 7. Specific power of rotating machines compared to turbine engine core
2) Cryogen storage
Weights of Various Cryocoolers and Subcomponents
Another possibility (which has not been extensively studied
yet) would be to load enough cryogen at the airport for the
100000
Table of Cryocooler Performance and weight.xls
flight duration including a margin; the cryo-tank would have
to be refilled after each landing. This may lead to a minimum
10000
weight associated with superconductor cooling. If liquid
hydrogen is the cryogen, it could also be used as fuel and
burnt in the engine along with jet fuel after cooling the HTS
Weight, lb

1000
components (e.g., 95% jet fuel, 5% H2).
3) LH2 as fuel
0.83
GM Cold Head lb = 8.0 hp
If LH2 is available onboard as fuel, either for turbines or
100
fuel cells, then cooling of the superconducting machines is
APPROX. RANGE OF “free” as we need to warm-up the hydrogen before being used
CURRENT INTEREST
as fuel. An excellent synergy is obtained and fully
10
superconducting machines are usable with very liberal limits
0.1 1 10 100 1000 10000
Input Power, hp on the AC losses created. However, dense hydrogen storage is
an issue; we may not see LH2-fueled aircraft in the N+3 time
Fig. 8. Specific power of commercially available cryocoolers frame (2030-2035), even though extensive studies of such
aircraft powered by conventional engines have been
Fig. 7 shows a comparison of the weight of electrical performed [4].
machines compared to turbine engine cores as a function of
power. As stated, conventional motors are too heavy to even IV. PAST AND CURRENT DEVELOPMENT OF
be considered. Current-day superconducting machines (with SUPERCONDUCTING MACHINES FOR AIRBORNE APPLICATIONS
resistive armature) are comparable with turbine engines in
Application of superconductors in propulsion was first
power density, while fully superconducting machines have the
potential to be 3 times lighter. HTS can be the enabling introduced in the context of ship propulsion [5], [6]. The first
 ESNF, No. 6, October 2008; ASC Preprint 2AP01 conforming to IEEE Policy on Electronic Dissemination, Section 8.1.9

mention of this application for aircraft was done by Oberly in Superconducting generators have already been
1976 at the Applied Superconductivity Conference [7]. The demonstrated to exhibit power densities in the range of turbine
main thrust of the Air Force program since then has been for engines thus validating the feasibility of future ultra
high power density generators in the multi-MW range to lightweight machines for airborne applications.
power weapons or onboard equipment [8]-[10], very little
B. Superconducting Motors
research has been done until recently on the use of
superconducting motors for electric propulsion of aircraft. The case for superconducting electric propulsion of aircraft
has been examined in [11]-[17]. In the concept of turbo-
A. Superconducting Generators electric propulsion presented in Fig. 5, not only ultra-compact
Having windings rotate above 10,000 RPM is very generators are needed, but also motors to power the propulsion
challenging as large acceleration forces are applied on the fans. Motors for propulsion application represent a much more
conductors. One machine under development by LEI is a challenging application as rotation speeds are lower than
3MVA/15,000 RPM generator pictured in Fig. 9. Both the generators and therefore electromagnetic torque needs to be
stator and rotor are operating at cryogenic temperature, but higher. This application was recently investigated as part of
only the excitation coils are superconducting; resistive losses our URETI program, results and design examples are
in cryogenic copper being lower than AC losses in HTS presented in next section.
conductors at the high operating frequency.
Another development involves a generator with the same V. NASA-DOD URETI’S SUPERCONDUCTING MACHINES FOR
requirements as the previous example but based on a different AIRBORNE APPLICATIONS
configuration. To address the high rotation speed, General This section briefly reviews our work as part of a 5-year
Electric used a bulk piece of magnetic material at the rotor research project sponsored by NASA and DoD: the University
magnetized by a stationary superconducting coil. This Research Engineering and Technology Institute (URETI) on
configuration provides a very robust rotor able to spin at high Aeropropulsion and Power [12], [14]. The main objective of
RPM. The flux distribution is not optimal but the high rotation this program was to investigate the feasibility of more-electric
speed brings the power density to an impressive 7 kW/kg. A and all-electric airborne vehicles, and in particular, the
rendering of the machine is shown in Fig. 10. possibility of incorporating superconducting machines into
electric propulsion schemes for future aircraft. Multiple
applications and vehicle configurations were investigated. The
first case study is a propeller-driven Cessna 172 type aircraft
chosen as a low power application [16]. Increasing power
density of small machines is very challenging as they present
very little room for coil winding. The first task was then to
develop a non-conventional motor topology applicable to low-
power machines.

Stator iron yoke

Stator windings

EM shield

HTS pancakes

Shaft HTS plates

Fig. 9. LEI/AFRL multi-MW superconducting generator
Bearings
Armature winding
Housing

Stator yoke
Fig. 11. Diagram of the superconducting motor

Solid rotor with A. General aviation aircraft

offset poles
For small aircraft, motors based on trapped flux may be
applicable. Design requirements were those of a Cessna 172:
120 kW of power at 2700 RPM. Because of the low power
level, a novel motor configuration was developed. A diagram
Stationary HTS coil (red) of the motor is shown in Fig. 11. The machine is composed of
Fig. 10. Homopolar Inductor Alternator from GE/AFRL Bi2223 pancake coils placed on the same axis and fed with
opposite currents. YBCO plates are evenly distributed around
ESNF, No. 6, October 2008; ASC Preprint 2AP01 conforming to IEEE Policy on Electronic Dissemination, Section 8.1.9

the axis and are used to trapped magnetic flux and concentrate equilibrate with the rest of the machine, thus trapping the
flux lines. Since the machine is expected to be conduction applied field. The current is then ramped down and reversed,
cooled, the superconducting inductor is stationary and the the YBCO plates will keep their magnetic flux constant by
outside copper air gap armature is rotating. The machine can means of undamped induced currents and the reversed flux
be small enough to fit inside a propeller hub, allowing the from the pancake coils is then concentrated between the plates
blades to be directly attached to its rotating armature, which giving a flux distribution such as that of Fig. 13, creating an 8-
collectively save the aircraft’s internal volume. pole machine. The machine operates synchronously and
exhibits the performance outlined in Table II.
1) Principle of operation
The presented topology is very promising and exhibits a
The generation of the eight pole excitation field is based on predicted power density in the same range as gas turbines
a 2-stage cooling system; the time evolution of the current in despite the low power level. The low power density of
the pancake coils is shown in Fig. 12 [17]. potential generation options (e.g., fuel cells) and cryocooler
overhead negates most weight gains and makes this concept
I field not practical yet.

I magnetization 2) Experimental validation

Since the feasibility of the proposed inductor is based on the
2-stage cooling system, a full-size thermal mock-up of the
inductor was constructed to experimentally validate the
Time cooling procedure [18]. The major challenge is to keep the
I operation YBCO Plates
cool down
YBCO plate above 90 K while the coils operate at full current
around 30 K. The large temperature gradient is handled by a 1
mm gap between the two parts placed in a vacuum cryostat.
Fig. 12. Time evolution of the current on the pancake coils. Fig. 14 shows the thermal mock-up before MLI wrapping.
Preliminary results obtained with a low power cryocooler
show that at least a 40K temperature difference between the
Bi2223 coils and the YBCO plates can be achieved in steady
state allowing the possibility of 2-stage cooling (Fig. 15).
Further experiments (with a larger cryocooler) will be done to
also validate steady-state gradients.

Temperature sensors
Fig. 13. Flux distribution in the machine. Bi2223 pancake coils

TABLE II YBCO plates

SUMMARY OF GENERAL AVIATION HTS MOTOR
Total length 160 mm
Fig. 14. Thermal mock-up of the inductor
External diameter 220 mm
Number of poles 8
Rotation speed 2700 RPM
Power 160 kW YBCO @
Total mass (including conduction cooling apparatus) 30 kg
Power density 5 kW/kg 100K
Heat load of superconducting part < 10W
Operating temperature 30 K

Adjacent
The whole system is cooled down to about 90 K when coil @
heaters are activated to maintain the YBCO plates above 60K
critical temperature, but the rest of the motor (including the
coils) continue to cool down to about 30K. The current in the
coils is then ramped up and a radial flux penetrates the YBCO
plates (which are non-superconducting at this time). Once the
Fig. 15. Experimental validation of the 2-stage cooling system
current achieved its nominal value, the heaters are shut down
and the YBCO plates are allowed to cool down and thermally
 ESNF, No. 6, October 2008; ASC Preprint 2AP01 conforming to IEEE Policy on Electronic Dissemination, Section 8.1.9

B. Safety Torque Generation Iron shield

On a single-engine airplane of course the issue of safety and Insulation layer

reliability is paramount, and we studied design options in case Superconducting coil

of a superconductor quench or loss of cooling [19]. Of the Stator support

options studied, the most promising is pictured in Fig. 16, Stator coils

consisting of adding a squirrel cage to the inductor so that the Bulk HTS plates

motor reverts from a synchronous motor to an induction Rotor support

(Displayed as wireframe)
machine in case of failure. Shaft
In case of a quench (failure) the propulsion motor does not Cryostat
have to provide full power, the airplane is able to maneuver to (Displayed as wireframe)

the nearest airport and land on only a fraction of the power

Fig. 17. Axial flux configuration for HALE aircraft
(typically 35-45%). The concept of Fig. 16 allows to generate
about 35% of full power with a decrease in overall power
density of about 35%. E. HTS Machines for Aircraft: Summary Conclusions
The results from the URETI work demonstrate that
superconducting motors and generators can be designed using
today’s materials to match the power density of turbine
engines. As Fig. 18 shows, the superconducting machines
considered in the URETI program, as well as others, can reach
much higher power densities than conventional machines,
even at relatively low power levels.
This is a very promising result that opens the possibility of
superconducting aeropropulsion. However, to fully replace
turbofans, the power density of the electrical components
needs to be even higher, which can only be achieved with
fully superconducting machines (inductor and armature). The
next section describes a physics-based model developed to
Fig. 16. Squirrel cage concept to provide safety torque in case of failure predict and optimize the weight of fully superconducting
machines as way of projecting what power densities are
C. Small Engine Study needed to achieve airworthiness.
Power density (kW/kg)
The second case study was to see if the motor topology 0 1 2 3 4 5 6 7 8

described above could be scaled up to a larger engine. The Oswald TF62 536kW @110RPM Conventional
Oswald TF46 506kW @200RPM
turbine engines in a typical small business jet are about Oswald TF36 326kW @300RPM Gas turbine cores
Torque
1.5 MW. The concept described above is modular, and more Oswald TF26 210 kW @400RPM optimized
Oswald TF20 63kW @500RPM commercial
HTS coils/YBCO plates can be stacked axially to increase Oswald TF13 28kW @500RPM

power. The power density of this system was estimated to be Conventional 4 MVA @3600RPM
URETI cylindrical 1.5 MW @3000RPM HTS
6.6 kW/kg, comparable to that of state-of-the-art turbines [20]. URETI axial flux 450 kW @3000RPM
Designed for
airborne
URETI cylindrical 170 kW @2700RPM
applications
D. High Altitude Long Endurance Aircraft GE HIA 5MVA @16000RPM
Siemens 4 MVA @3600RPM
The final case study is concerned with a High-Altitude, Siemens 400 kW @1500RPM Actual HTS
motors
Long-Endurance (HALE) air vehicle (e.g., a hurricane AMSC 3.7 MW @1800RPM
AMSC 5 MW @230RPM
tracker). This would be an unmanned aircraft, fully electric, AMSC 36.5 MW @120RPM

able to fly and loiter for up to 14 days without refueling or Torque density (Nm/kg) 0 5 10 15 20 25 30 35 40
Power density (kW/kg) Torque density (Nm/kg)
returning to base. For maximum efficiency, the
superconducting motor for the propulsor needs to be both Fig. 18. Comparison of concept designs and existing HTS machines with
extremely light and compact, but also have very low losses. respect to conventional motors and turbine engines in terms of power density
We chose a lead-less axial flux configuration (allowing for
higher trapped flux for compactness). The design concept, VI. PHYSICS-BASED SIZING MODELS FOR THE MORE/ALL-
described in [21] and shown in Fig. 17, is projected to achieve ELECTRIC AIRCRAFT DESIGN
an impressive power density of 7.4 kW/kg using conventional Of particular importance in the URETI program was the
HTS materials available today. development of physics-based models for superconducting
machines for integration into existing aircraft design tools
(i.e., realistic size and weight models for superconducting
components to use in aircraft system design studies).
 ESNF, No. 6, October 2008; ASC Preprint 2AP01 conforming to IEEE Policy on Electronic Dissemination, Section 8.1.9

A. Aircraft Design Based on New Propulsion Technology B. Fully Superconducting Machine Sizing Model
Integrating a single disciplinary technology into aircraft An integrated electromagnetic/thermal model of
systems often results in an array of cross-disciplinary effects. superconducting rotating machines has been developed to be
This is particularly true in the case of the HTS machinery integrated into aeropropulsion system design/analysis tools
technology, because it will entail fundamental changes in not developed at Georgia Institute of Technology and NASA
only the propulsion system but also the subsystem architecture Glenn Research Center [31]. The model is composed of an
and airframe design. Such complex nature of aircraft systems analytical electromagnetic model and a lump-parameter
integration may disqualify a premature technology evaluation thermal model [32]. The model architecture is shown in Fig.
made solely relying upon its component level metrics such as 20.
power density and efficiencies [22], [23].
best states weight
In addition, the introduction of the HTS motor drive power GLOBAL OPTIMIZATION 1500
1250
system will usher in an uncharted design space that POWER
RPM ELECTROMAGNETIC
1000
750
encompasses revolutionary propulsion system architectures, MATERIAL MODEL
500
250

alternative energy sources and storage, and a myriad of 20 40 60 80 100

k

OPTIMIZATION
options to integrate them into an airframe. For instance, the THERMAL
MODEL MINIMUM WEIGHT
CONSTRAINTS OR VOLUME
generators and gas turbine engines in Fig. 5b can be
potentially replaced with any electric power source such as
fuel cells, high performance electric batteries, or ultra Fig. 20. Block diagram of a fully integrated electromagnetic-thermal model of
capacitors. Although they are not readily applicable to high a superconducting machine tied to an optimizer to achieve minimum volume
or weight
power systems due to their low power and/or energy density,
those technologies, providing a supplemental power, may 1) Machine Configurations
serve to facilitate the integration of a HTS motor drive Several machine configurations are possible such as radial
propulsion system. flux machines or trapped flux magnet excited machines. For
Therefore, the true value of airborne HTS applications need this study, we will limit the configuration of the
be evaluated at the air-vehicle integration level at least in light superconducting machines to radial flux and distributed
of the benefits from a synergistic integration of facilitating windings. The windings are usually made of racetrack coils
technologies. To this end, various combinations of aircraft that will be considered as continuous current distributions in a
configurations, technologies, and missions must be assessed in first approximation. Flux density in the backiron is limited to
terms of vehicle-level metrics such as aircraft weight, field 1.7T and the J(B) operating point of the conductor is
length, fuel burn, emissions, and noise through an aircraft calculated.
sizing and synthesis process at the appropriate level of fidelity
and engineering realism. Nevertheless, such revolutionary
Back iron
concepts are very likely to substantially deviate from a HTS field winding
historical trend, thereby necessitating the development of Stator winding

physics-based analysis capabilities (illustrated in Fig. 19).

Electromagnetic shield

Physics-Based Modeling
and Simulation Sensitivity Analysis
Notional Design
Ind uced 2.72972
Technologies Impact of
CD0

2.417712
Drag ±0.0014
2.09899

0.00053
Parasite0.000237
Technologies es eis
r1
K

Probabilistic Drag±6.6e-6
0.00014
Risk
r2
and Statistical
1

1
-1

-1

-1

-1

0 0 0 0

AR
AR Taper Wing Sweep
Taper Ratio Sw eep Angle Wing Area Uncertainty
Techniques Ratio Area ro

rs

Fig. 19. Block diagram representing a physics-based model approach to re

aircraft design and optimization
Fig. 21. Generic superconducting machine configuration
Addressing this deficit, a joint effort [24], [25] under the
URETI program has indentified and developed key enabling 2) Electromagnetic Model
capabilities for electric aeropropulsion and aircraft modeling For this particular model only Bi2223 and YBCO
and simulation, including revolutionary propulsion conductors are considered, their properties have been
architecture modeling focusing on fuel cells [26], energy- determined from data available in literature. The following
based aircraft sizing and synthesis [27], and volumetric sizing equation shown the relation used to represent the YBCO
methods [28]. The addition of probabilistic methods such as coated conductors and Fig. 22 shows the characteristic of
reliability-based design optimization [29], [30] and stage- Bi2223 considered for the simulations.
based recourse programming [26] allows the effect of −
B
 T 
uncertainty to be embraced, resulting in a reliable and robust J (0 T ,77 K )  T  90  1− 
(1)
solution for rapidly advancing technologies. J e ( B, T ) = e 1 −  e  TB 
77  Tc 
1−
Tc
ESNF, No. 6, October 2008; ASC Preprint 2AP01 conforming to IEEE Policy on Electronic Dissemination, Section 8.1.9

With Tc =92 K and TB =80 K 8 (6)

PAC ≈ J c d f Br0 Vsc
3π
where J c is the critical current density of the conductor, d
6 1,2 Jc/Jco (//ab ; 0 T)

5 1
Jc/Jco (//ab ; 1 T)
Jc/Jco (//ab ; 2 T)
the dimension of superconductor that faces the flux (filament
PIT tape

Jc (T, B)/Jc (20 K, B)

4
20 K
// ab planes 0,8
Jc/Jco (//ab ; 3 T) size for Bi2223), f the frequency and Vsc the volume of
Fit
superconductor.
co
J /J

3 0,6
c

2 0,4 4) Thermal Model

1 0,2 For both superconducting and conventional machines, the
0 0 limit of power output comes from the amount of heat
0 2 4 6 8 10 10 20 30 40 50 60 70 80
B (T) Temperature (K)
generated in the conductors. In the case of conventional
superconducting machines where only the rotor is
Fig. 22. Critical current density of Bi2223 tapes (from AMSC) superconducting, the limiting factors are the critical current
curve for the superconductor, and the maximum allowed
The sizing of the machine is performed based on power and temperature of the armature windings and thus the amount of
speed requirements; from speed Ω and power P, the Joule heating in the copper. Therefore, accurate estimation of
electromagnetic torque T can be obtained using equation 2. the maximum temperature of the armature winding is an
P important part of the machine design. Such calculation can be
T= (2)
Ω done using an equivalent lumped parameter circuit model as
From the electromagnetic torque, the no-load field Br0 can shown in Fig. 23.
For a fully superconducting machine, the AC losses in the
be determined as follows:
armature winding need to be removed by the cooling system.
T (3)
Br0 = Depending on the cooling available, thermal transients can
 L  occur. An equivalent lumped parameter model including AC
2 K S π r03  a 
 r0  losses can be used to simulate the thermal transient response
With K S the electrical loading defined in equation 4, r0 the of the superconducting armature winding and assess the
minimum amount of cooling power required for stable
mean armature radius (cf. Fig. 21), and La / r0 the aspect ratio
operation.
of the machine. Tamb
Ueca
3N s kd I s (4)
Reca
Ks =
π r0 Ieca
Tc

The thickness of the armature winding can be determined Uc,2 Rc,2

Ie Iw
from the value of K s and the thickness of the rotor windings Ic2
Stator
case
TFe
can be obtained from the value of Br0 using the following Ria,ec Uia,ec
Uc,1
PFe
Rc,1
equation. Uew,ec

β
  
Rew,ec
J f r2 sin p +1 p+ 2 2p
0
B =
2µ 0 2  r2  1 −  r1   1 +  r0   (5) Armature
Rg Ug Air gap

p + 2  r0    r2     rs  
r winding
π I1
  
end parts
Rew,ia Uew,ia Armature
where Jf is the rotor current density, β is the winding aperture, R2 U2 winding
Ucv
p the number of pair of poles and µ 0 the permeability of TCu Icv

vacuum. The weight and volume of the machine can then be PCu
R1
Rcv

computed together with the cryostat, torque tubes, mechanical U1

structure, shaft, etc. The synchronous reactance that gives Is,ag

Us,ag

important information about the dynamic behavior of the Rs,ag

machine can also be calculated.

3) AC Losses Fig. 23. Example of lumped parameter thermal model used to calculate the
resistive armature maximum temperature
For a fully superconducting machine, AC losses represent
the major part of the heat load. It is important to have a good 5) Optimization
approximation of the AC losses. They can be separated into For a given power (or torque), an optimization method is
three categories: the magnetization losses, transport current used to minimize the weight or volume based on the coupled
losses and the coupling losses in the matrix. The dominance of electromagnetic and thermal models (using appropriate
one type of losses over the others depends on the frequency of physical constraints) to make sure that the output sizing is
current and applied field. In a first approximation, we have realistic. Simulated annealing method is used to avoid local
limited the losses computation to magnetization losses as we minima and cover the entire design space.
think they will be dominant in a rotating machine application.
They can be calculated using equation 6.
ESNF, No. 6, October 2008; ASC Preprint 2AP01 conforming to IEEE Policy on Electronic Dissemination, Section 8.1.9

10

VII. AIRCRAFT DESIGN EXAMPLES and, as a result, an off-set inlet is required to duct the intake
air to the engine. Electric cables transfer energy to the fan
A. Application of Superconducting Machine Models to
Aircraft Design motors, and there would be cross cabling to the other side of
the aircraft to complete an electrical bus so that all motors can
Electric propulsion systems with superconducting draw from both generators.
technology can potentially be applied to any future aircraft,
manned or unmanned. But to have the maximum impact on
the critical issues of the environment, electric propulsion must
be introduced into subsonic transport aircraft designs of the
future. Two potential transport aircraft with electric propulsion
are discussed in this section: 1) a small regional transport
aircraft with a design having a conventional arrangement of
fuselage and high aspect ratio wing, and 2) a large
transcontinental or intercontinental transport having hybrid-
wing-body (HWB) design and short takeoff and landing
(STOL) capabilities. For both designs, electric power
production is obtained with turbine engines driving
superconducting generators, and propulsive thrust is
distributed on the lifting surfaces with multiple electric motors
driving single-stage fans.
Fig. 24. Super-Conducting Electric Propulsion
B. Small Regional Jet
Noise from the buried turboshaft engine will easily be
As described earlier, NASA has defined several future time-
attenuated, and the jet noise from these engines is essentially
frames in which technology measured by noise level,
non-existent. Tip speed for the fans is low and this noise from
emissions and mission fuel burn is successively reduced from
the fans is partially shielded from the ground by the wings.
levels of current technology (see Table I). A study is now
Although a full noise study for this aircraft has not been
being conducted to design short-field regional subsonic
completed, it is expected to be very quiet and meet the N+2
transport aircraft having a full payload of nominally 100
targets of Table I.
passengers [33]. These aircraft are for the N+2 time frame,
The engine design is patterned after that of the General
and the study has been extended to include a design having a
Electric T700 turboshaft engine with a single spool
superconducting electric propulsion system (for possible N+3
compressor having multiple axial stages followed by a single
introduction).
centrifugal stage. The major technology in this propulsion
In this design, two advanced turboshaft engines are buried
system is obviously with the superconducting generator. The
within the lower fuselage each with direct drive to an electric
generator is designed using the methodology outlined in this
generator. Electric power is transmitted to multiple motors
paper, and the result is truly remarkable. The diameter of the
each with direct drive to a single stage fan (see Fig. 24). Both
generator at 10.24 inches is half that of the maximum engine
generators and motors are assumed to be fully
diameter, and the light weight of the fully superconducting
superconducting. The fan-motor sets are mounted above and
generator yields a power to weight ratio of 40 HP/lb
toward the wing trailing edge in individual nacelles with a
(66 kW/kg). The generator rotates at engine rotational speed
total of ten (10) sets – five per wing. The turboshaft engines
resulting in reduced torque and very light weight (335 lb each
are relatively near term and the development of the fans will
generator, with each turbine engine at 894 lb).
not require a major new development although they may
Five fans per wing are installed above the wing with the
require either variable pitch blades or a variable area nozzle.
exhaust nozzle near the trailing edge. This nozzle is a two-
Thus it is the development of the superconducting generators
dimensional variable area design to match energy and flow
and motors that will enable and pace the future introduction of
requirements at both takeoff and cruise. Light weight, wide
such an aircraft. It should be noted that electric propulsion
chord composite fan blades are assumed using the technology
offers a level of safety in the event of the loss of an engine not
of current high bypass ratio turbofan engines. Advanced
available in current transport aircraft. Albeit at reduced power,
technology is assumed in the low value of hub-tip ratio at the
all ten fan-motor sets can be run with a single turboshaft
entrance of the fans and high inlet Mach number at the fan
engine. In fact, this may be a desirable mode of routine
entrance annulus. These parameters are selected to reduce fan
operation in some segments of the standard aircraft mission
diameter.
such as descent and landing.
As with the generators, major new technology is applied to
The engine, generator, motors and fans shown in the figure
the superconducting motors. The fully superconducting motor
are to scale, and, as can be seen, the integration of the
outside diameter at 7.24 inches is an excellent match with the
propulsion system into the aircraft does not pose any
hub diameter of the fan exit, and the light weight of the motors
significant problems. The generator is placed in front of the
is based on a power to weight ratio of 24.6 HP/lb (40 kW/kg),
engine to avoid having to insulate it from the hot exhaust gas,
a lower power density that the generators. The motors rotate at
ESNF, No. 6, October 2008; ASC Preprint 2AP01 conforming to IEEE Policy on Electronic Dissemination, Section 8.1.9
11

fan rotational speed, and the torque-power ratio is somewhat C. Distributed Turboelectric Propulsion for Hybrid Wing
higher in this case. Nonetheless, these light weight motors Body Aircraft
contribute to a very light weight propulsion system. Each Meeting future goals for aircraft and air traffic system
motor weighs 110 lb, and with cables included, the total performance may well require new airframes with more highly
turboelectric propulsion system weighs slightly more than integrated propulsion. Previous studies have evaluated hybrid
5100 lbs. wing body (HWB) configurations with various numbers of
The NASA aircraft design study of [33] is continuing, and a engines and with increasing degrees of propulsion-airframe
preliminary comparison of the aircraft described in this section integration. One recently published configuration [11] with 12
has been made with a similar aircraft designed with high small conventional engines partially embedded in a HWB
bypass ratio turbofan engines. These turbofan engines are aircraft (shown in Fig. 25) served as the airframe baseline for
assumed to be an advanced version of the new geared turbofan the turboelectric concept aircraft described below.
(GTF) engine now under development by Pratt and Whitney To achieve high cruise efficiency, this high lift-to-drag ratio
Aircraft. The design point bypass ratio of the engine is HWB was adopted as the baseline airframe along with
nominally 10 with a fan pressure ratio in the 1.45-1.5 range. boundary layer ingestion inlets and distributed thrust nozzles
The engines for this aircraft are placed above the wing to to fill in the wakes generated by the vehicle. The distributed
provide shielding to reduce noise. One can consider that the powered-lift propulsion concept for the baseline vehicle used a
electric propulsion system also represents a very high bypass simple, high-lift-capable internally blown flap or jet flap
ratio engine with the fan (bypass) flow displaced from the core system with a number of small high bypass ratio turbofan
flow through the turboshaft engines. In the design presented engines in the airframe. In that concept, the engine flow path
above, the equivalent bypass ratio is approximately 13. from the inlet to the nozzle is direct and does not involve
A preliminary comparison of the two aircraft is given in complicated internal ducts through the airframe to redistribute
Table III. Both aircraft are designed for short-field operation, the engine flow. In addition, partially embedded engines,
and, as a result, they have relatively high total takeoff thrust to distributed along the upper surface of the HWB airframe,
gross weight ratios. To accommodate the multiple fans, the provide noise reduction through airframe shielding and
aircraft with the electric propulsion system is designed with a promote jet flow mixing with the ambient airflow.
larger and higher aspect ratio wing. The result is a much
shorter field length capability for this aircraft. Range (nm) 3,000
The gross weight of the electric powered aircraft is Payload (lb) 40,000
approximately 5% lower than the turbofan powered aircraft Takeoff Gross Weight (lb) 189,000
primarily due to a reduction in the propulsion system weight, Landing Weight (lb) 152,000
but no comparison of the range of each aircraft has been made Total Fuel (lb) 44,100
yet. However, the electric powered aircraft does have higher Block Fuel (lb) 37,700
parasite drag due to the added surface area of the fan nacelles,
Block Time (h) 6.92
and this will have an effect on cruise aerodynamics and thus
Initial Cruise Altitude (ft) 39,000
on range capability. Also, note that the thrust specific fuel
Takeoff Field Length (ft) 2,450
consumption of the electric propulsion system is slightly
greater than that of the high bypass ratio turbofan propulsion
system. With the proper engine cycle for the turboshaft
engine, it is expected that the specific fuel consumption at Fig. 25. Hybrid Wing Body Aircraft used as baseline airframe for
cruise will be comparable for both aircraft. A complete turboelectric HWB.
evaluation of range will include operation of the aircraft. It is
anticipated that aircraft with an electric propulsion system will
be able to operate with a single engine operating at high
efficiency through the descent phase of the mission thus
offering further fuel savings.

TABLE III
COMPARISON OF AIRCRAFT PERFORMANCE
N+2 Characteristic N+3
0.80 Cruise Speed (Mach) 0.80
36,000 Cruise Altitude (ft) 36,000
83.33 Wingspan (ft) 100
9.29 Aspect Ratio 10.36
81,650 Takeoff Gross Weight (lbs) 77,311
109 Takeoff Wing Loading (psf) 80
0.46 Takeoff Thrust-to-Weight Ratio 0.42 Fig. 26. Two views of a sixteen-fan hybrid wing body aircraft and a cross
4,644 Takeoff Field Length (ft) 2,415 section of one of the superconducting-motor-driven fans in its duct.
3,132 Landing Field Length (ft) 1,755
 ESNF, No. 6, October 2008; ASC Preprint 2AP01 conforming to IEEE Policy on Electronic Dissemination, Section 8.1.9
12

To improve performance and to reduce noise and and efficiency comparisons are made in Table IV among three
environmental impact even further, a turboelectric propulsion propulsion systems: a 16-fan turboelectric propulsion system,
system was proposed for that vehicle. The turboelectric 16 independent small turbofan engines, and 2 large
concept aircraft [13] shown in Fig. 26 uses essentially the conventional turbofans. The turboelectric system weighs 5000
same airframe but employs a number of superconducting lb (2300 kg) more than the 16-engine system but has 9% lower
motors to drive the distributed fans rather than many small TSFC (Thrust Specific Fuel Consumption) including the 1%
conventional turbofan engines. The power to drive the electric electrical and refrigeration loss at takeoff. Weights exclude
fans is generated by two remotely located gas-turbine-driven propulsors (fans), which would have similar total weights in
superconducting generators (at each wing-tip). This all systems. TSFC values shown in Table IV are based on best
arrangement allows many small partially embedded fans while present-day values for the engine size. Refrigerator weight is
retaining the superior efficiency of large core engines, which based on 5 lb/HP-input (3 kg/kW-input) and 30% efficiency,
are physically separated but connected through electric power and HTS AC losses corresponding to a 12-µm filament. As
lines to the fans. Descriptions of the vehicle, the noted above, even with this optimistic assumption (about
superconducting system, and the propulsion system are equal to the weight of just the cold head in today’s
described in [13] with some "zeroth-order" weight and cryocoolers), the weight overhead of the refrigeration is such
efficiency comparisons to the multiple turbofan system. that the optimizer pushes the operating temperature of the
Preliminary analysis suggests that fuel savings might be superconducting components to a relatively high 49K (at the
greater than six percent for a turboelectric propulsion system expense of having larger and heavier electric machines) to
compared to the same frame with distributed discrete reduce refrigeration needs.
turbofans. To compare the turboelectric system with the 16-turbofan
Beyond fuel savings related only to the propulsion system, system, by balancing out the opposite effects of lower SFC and
however, turboelectric propulsion introduces a very high higher weight of the turboelectric system, the Breguet range
degree of aircraft design and operational flexibility as a result equation, sufficient to determine relative ranking, is applied to
of decoupling power production from power consumption, as both systems, with the requirement of equal aircraft range and
has been noted earlier in this paper. Lightweight approximating the entire flight as cruise. Solving for the
superconducting generators, motors and power cables allow a required change in fuel weight between the 16-engine case and
small number of efficient large turbo-generators to power an the turboelectric case, we find that the turboelectric aircraft
arbitrary number of propulsor units. Either can be placed would require 7%, or 3000 lb (1400 kg) less mission fuel. Thus,
practically anywhere and in various orientations on the the slightly heavier turboelectric aircraft would have a net fuel
vehicle. This flexibility opens up design possibilities not savings of roughly 7% on each flight, compared to the baseline
obtainable with discrete large turbofans or with distributed aircraft powered by 16 small engines. This estimate must be
propulsion systems that employ mechanical power distribution refined by a detailed mission analysis. Known omissions in
by gearboxes and shafts. Fuel savings resulting from this the weight estimates of the electric system include the
design freedom may be book-kept as drag reduction rather superconducting transmission lines (estimated at only 3% of
than under thrust production. Remembering that thrust must the turboelectric system weight) and other power management
equal drag in steady level flight, we may further note that and distribution components.
there is only a limited amount of fuel saving to be gained by A comparison between the turboelectric case and two large
engine improvement whereas fuel saving from drag reduction (presumably podded) turbine engines can also be made based
is more open ended. Large engines already extract half or on the numbers in Table IV. One can see that the entire
more of the fuel energy, very close to the thermodynamic refrigerated turboelectric system weighs 6300 lb (2900 kg)
limit. Drag reduction by boundary layer ingestion and wake more than two large turbofan engine cores of 42,000 HP each
filling are yet to be fully analyzed and exploited and will (with no weight allowance for podding) and would be ~1%
require engine/airframe integration and a distributed less efficient at takeoff because of the electrical losses. A
propulsion approach to be realized. liquid-hydrogen-cooled turboelectric system would weigh
In spite of uncertainty of the future level of refrigerator and 3600 lb (1600 kg) more than the large turbofan engine cores.
AC superconductor technology, we summarize some weight Thus, the propulsion system weight for an HWB using podded
and efficiency estimates from [13] that are based on the level engines would be significantly less than either of the two
of development that we expect for all-superconducting turboelectric systems discussed, with consequent
generators and motors. A sizing code [34] for fully accompanying reductions in fuel burn. However, the use of
superconducting motors and generators was used. two separate podded engines would provide no short take-off
Optimization was performed to minimize motor (or generator) capability and only limited noise reduction, no drag reduction
weight plus refrigerator weight. The refrigerator, even with and none of the other potential benefits and capabilities that
our aggressive 2030 assumptions, weighs ~70% as much as have been mentioned above. It is therefore still very
the motor or generator that it cools. Efficiencies, including the advantageous to switch to turboelectric propulsion, even with
refrigerator power, are at least 99.4%. The expected weight of the slight weight penalty.
a motor or generator with its cooler is considerably less than
the weight of a turbine engine core for equal power. Weight
ESNF, No. 6, October 2008; ASC Preprint 2AP01 conforming to IEEE Policy on Electronic Dissemination, Section 8.1.9

13

TABLE IV much more energy efficient, thus enabling sustainable

COMPARISON OF DIFFERENT PROPULSION SYSTEMS aviation. Conventional electric machines are too heavy to ever
Propulsion Weight, lb Efficiency, TSFC,
be considered for this extremely weight sensitive application,
Components thereby warranting the need to investigate the feasibility of
System (kg) % hr–1
Two 42 380
7300 superconducting machines as an alternative: motors to drive
hp engine ---- 0.57 propulsion fans, and generators to power such load.
(3300)
cores
Two 42 380 The authors reviewed numerous case studies concerning the
hp electric
3000
application of superconducting rotating machines to aircraft.
generators 99.7 ---- Detailed design studies for HTS propulsion motors supported
(1300)
Turboelectric (including
distributed fans refrigerators) by experimental validation have convinced us that
(refrigerated) Sixteen superconducting rotating machines today can achieve power
5250-hp densities comparable with that of turbine engines (3-8 kW/kg).
4700
motors 99.4 ----
(including
(2100) This remarkable achievement, however, is still not enough for
refrigerator) deployment into commercial aircraft. Electrically-propelled
15 000 airliner aircraft would become feasible when power densities
Total 99.1 ----
(6800)
Two 42 080-
approach 25 kW/kg for motors and 50 kW/kg for generators,
7300 which appears to be achievable with fully superconducting
hp engine ---- 0.57
(3300)
cores machines (both inductor and armature).
Two 42 080- The design examples of HTS motor-drive aircraft that were
hp electric 1900
99.9+ ---- studied indicate that turbo-electric propulsion using
Turboelectric generators (860)
distributed fans (LH2 cooled) superconducting machines can substantially contribute
(LH2 cooled) Sixteen towards achieving the aggressive goals set for overall fuel
5250- hp 3100
motors (LH2 (1400)
99.9+ ---- efficiency. This is primarily due to the separation of power
cooled) generation devices and propulsors, which offers an
12 300 unprecedented level of design freedom facilitating the
Total 99.9 ----
(5600)
Conventional
integration of short take-off capabilities into aerodynamically
Sixteen efficient body shapes (i.e, very high lift/drag ratio).
small 10 000
5250-hp 91* 0.63
distributed
engine cores
(4500) This promising new application cannot come to fruition
turbofans
without further research and development on HTS
Conventional
Two 42 000- superconducting materials and refrigeration technology. A
large 8700
hp engine ---- 0.57
nondistributed
cores
(4000) development roadmap includes:
turbofans • Develop and demonstrate fully superconducting
* Relative to 42,000-HP engine core at 0.57 thrust specific fuel consumption.
rotating machines in the range of 25-40 kW/kg for
If the motors and generators were cooled by liquid motors, and 40-80 kW/kg for high rotation speed
hydrogen (with only enough carried on the aircraft to provide generators (up to 15,000 RPM)
refrigeration) rather than refrigerators, then the turboelectric • Develop low AC loss HTS conductors (<10 W/A-
system would weigh 2300 lb (1000 kg) more than the 16- m @ 500Hz, equivalent to 10 µm filament) for
engine system, and the required jet fuel is reduced by 4000 lb fully superconducting machines
(1800 kg), or 9% (calculated from the efficiency advantage of • Develop cryocoolers capable of 30% of carnot
the large engines, without accounting for the replacement of efficiency and weighing less than 3 kg/kW-input
jet fuel energy with liquid hydrogen energy), and TOGW (or alternative lightweight refrigeration schemes)
(Take-Off Gross Weight) drops by 560 lb (255 kg). This • Refine the physics-based models for
estimate does not include corrections for the weight of the superconducting machines and ancillaries to
liquid hydrogen (which would provide about 5% of the continue exploration of aircraft design space and
aircraft’s fuel energy) and its tankage and accessories, alternative concepts
compared to the corresponding weight reduction of the jet It is important to consider that safety and reliability are
fuel, tankage, and components. (It may be noted that, for the primary considerations in the aircraft industry, so that no new
same energy, liquid hydrogen has almost 4 times the volume technology will be introduced without a solid track record of
but only one-third the weight of jet fuel.) implementation and test. Even though the concepts discussed
here may be slated for first service in 20-25 years, the
VIII. CONCLUSIONS AND RESEARCH DIRECTIONS superconducting machines to power them need to be a
Electric aircraft have long been considered due to a number somewhat mature and tested technology 15-20 years from
of operational and environmental benefits that could be now. This is an instance in which the market pull will be
derived from such transition. Electric aeropropulsion would strong, but the technology has to be ready before the pull is
offer tremendous benefits in the design of aircraft, ushering in fully expressed. It means that funding and effort for the
the possibility of revolutionary concepts that are quieter and developments outlined above need to start soon, and be
ESNF, No. 6, October 2008; ASC Preprint 2AP01 conforming to IEEE Policy on Electronic Dissemination, Section 8.1.9

14

sustained for the next decade or more. It should be pointed out Transactions on Applied Superconductivity, Vol. 15, No. 2, pp 2150-
2153, June 2005
that the development of ultra-compact superconducting [19] P.J. Masson, P. Tixador and C.A. Luongo, “Safety Torque Generation in
generators and motors, and the associated lightweight HTS Propulsion Motor for General Aviation Aircraft,” IEEE.
cryocooling, could also be applied to other mass-market Transactions on Applied Superconductivity, Vol. 17, No 2, pp 1619-
transportation applications such as trucks, busses, or even cars 1622, June 2007
[20] P.J. Masson, J.E. Pienkos and C.A. Luongo, “Scaling Up of HTS Motor
if compact enough. Realization of these applications could be Based on Trapped Flux and Flux Concentration for Large Aircraft
the next big opportunity for superconductivity to fully realize Propulsion,” IEEE Transactions on Applied Superconductivity, Vol. 17,
its game-changing potential. No 2, pp 1579-1582, June 2007
[21] P.J. Masson, M. Breschi, P. Tixador, and C.A. Luongo, “Design of HTS
Axial Flux Motor for Aircraft Propulsion,” IEEE Transactions on
ACKNOWLEDGMENT Applied Superconductivity, Vol. 17, No 2, pp 1533-1536, June 2007
[22] D.P. Raymer, Aircraft Design, A Conceptual Approach (3rd Edition),
The authors thank Dr. Danielle Soban of ASDL/Georgia American Institute of Aeronautics and Astronautics, Reston, 1999.
Tech for the helpful discussions on aircraft design, and Prof. [23] J.D. Anderson,. Aircraft Performance and Design, McGraw-Hill,
Pascal Tixador of Grenoble INP for his work on the sizing Boston, 1998
[24] T. Choi, T. Nam, and D. Soban, “Utilizing Novel Synthesis and Analysis
models. Methods towards the Design of Revolutionary Electric Propulsion and
Aircraft Architectures,” AIAA-2005-7188, Sep. 2005
[25] D.S. Soban and E.G. Upton, “Towards Electric Aircraft: Progress under
the NASA URETI for Aeropropulsion Power Technology,” SAE 2006-
REFERENCES 01-3097, Nov. 2006
[1] N. Cumpsty, Jet Propulsion, Cambridge University Press, 2002 [26] T. Choi, “A recourse-based solution approach to the design of fuel cell
[2] D. Blanding, "Subsystem Design and Integration for the More Electric aeropropulsion systems,” Ph.D thesis, Georgia Institute of Technology,
Aircraft," AIAA-2007-4828, 2007 2008
[3] H.J.M. ter Brake, and G.F.M. Wiegerinck, “Low-power cryocooler [27] T. Nam, “A Generalized Sizing Method for Revolutionary Concepts
survey,” Cryogenics, vol. 42, pp. 705-718, Nov. 2002 under Probabilistic Design Constraints,” Ph.D thesis, Georgia Institute
[4] G. Daniel Brewer, Hydrogen Aircraft Technology, CRC Press, Boca of Technology, 2007
Raton, FL, 1991. [28] E.G. Upton, “An Intelligent, Robust Approach to Volumetric Aircraft
[5] W.J. Levedahl, “Superconductive Naval Propulsion Systems,” Sizing,” Ph.D thesis, Georgia Institute of Technology, 2007
Proceedings of the Applied Superconductivity Conference 1972, May 1- [29] T. Nam, D.S. Soban, and D.N. Mavris, “A Non-Deterministic Aircraft
3, Annapolis, MD., USA, IEEE Pub. No. 72CHO682-5-TABSC, pp. 26- Sizing Method under Probabilistic Design Constraints,” AIAA-2006-
32, IEEE, 1972 2062, May 2006
[6] G.R.Fox and B.D. Hatch “Superconductive Ship Propulsion Systems,” [30] T. Nam and D.N. Mavris, “Multi-Stage Reliability-Based Design
ibid., pp. 33-40 Optimization for Aerospace System Conceptual Design,” AIAA-2008-
[7] C. E. Oberly, “Air Force Applications of Lightweight Superconducting 2073 Apr. 2008
Machinery”, IEEE Trans. Mag., vol. 13, No. 2, pp 260, January 1977 [31] NASA, Numerical Propulsion System Simulation (NPSS) User Guide,
[8] S.S. Kalsi, K. Weeber, H. Takesue, C. Lewis, H.W. Neumueller, and Rev. 1.6.4Q. Cleveland, OH, November 5, 2006
R.D. Blaugher, “Development Status of Rotating Machines Employing [32] P.J. Masson, P. Tixador, J.C. Ordoñez, A. Morega and C.A. Luongo,
Superconducting Field Windings,” Proc. IEEE, vol. 92, No. 10, pp. “Electro-Thermal Sizing Model for HTS Motor Design,” IEEE
1688–1704, Oct. 2004 Transactions on Applied Superconductivity, Vol. 17, No 2, pp 1529-
[9] P.N. Barnes, M.D. Sumption, and G.L. Rhoads, “Review of High Power 1532, June 2007
Density Superconducting Generators: Present State and Prospects for [33] “The Integrated Modeling and Verification of Hybrid Wing-Body, Low
Incorporating YBCO Windings,” Cryogenics, Vol. 45, No. 10–11, pp. Noise ESTOL Aircraft,” NASA NRA Contract with California State
670–686, Oct.–Nov. 2005 Polytechnic University, San Luis Obispo, CA., 2008
[10] C. Oberly, “Lightweight Superconducting Generators for Mobile [34] P.J. Masson, and P. Tixador, Superconducting machines sizing model,
Military Platforms,” Proceedings of the PES Meeting, Montreal, private communication.
Quebec, June 2006
[11] H.D. Kim, J.J. Berton, and S.M. Jones, “Low Noise Cruise Efficient
Short Take-Off and Landing Transport Vehicle Study,” AIAA–2006–
7738, Sept. 2006.
[12] P.J. Masson, D.S. Soban, G.V. Brown and C.A. Luongo, "HTS
Machines as Enabling Technology for All-Electric Airborne Vehicles,"
Superconductor Science and Technology, Vol 20, No 8, pp 748-756,
August 2007
[13] H.D. Kim, G.V. Brown, and J.L. Felder, “Distributed Turboelectric
Propulsion for Hybrid Wing Body Aircraft,” 9th International Powered
Lift Conference, London, United Kingdom, July 2008.
[14] P.J. Masson and C.A. Luongo, "HTS Machines for Applications in All-
Electric Aircraft," 1-4244-1298-6/07, Proceedings of the PES Meeting,
Tampa, FL, 2007
[15] G.V. Brown, A.F. Kascak, B. Ebihara, D. Johnson, B. Choi, M. Siebert,
and C. Buccieri, “NASA Glenn Research Center Program in High Power
Density Motors for Aeropropulsion,” NASA/TM—2005-213800, 2005
[16] P.J. Masson, D.S. Soban, E. Upton, J.E. Pienkos and C.A. Luongo,
“HTS Motors in Aircraft Propulsion: Design Considerations,” IEEE.
Transactions on Applied Superconductivity, Vol. 15, No. 2, pp 2218-
2221, June 2005
[17] P.J. Masson and C.A. Luongo, “High Power Density Superconducting
Motor for All-Electric Aircraft Propulsion,” IEEE Transactions on
Applied Superconductivity, Vol. 15, No. 2, pp 2226-2229, June 2005
[18] J.E. Pienkos, P.J. Masson, S.V. Pamidi and C.A. Luongo, “Conduction
Cooling of a Compact HTS Motor for Aeropropulsion,” IEEE.