-
Original Article
A methodology for vehicle and missionlevel comparison of More
Electric Aircraftsubsystem solutions: Application to theflight
control actuation system
Imon Chakraborty1, Dimitri N Mavris1, Mathias Emeneth2
andAlexander Schneegans2
Abstract
As part of the More Electric Initiative, there is a significant
interest in designing energy-optimized More Electric Aircraft,
where electric power meets all non-propulsive power
requirements. To achieve this goal, the aircraft subsystems must
be
analyzed much earlier than in the traditional design process.
This means that the designer must be able to compare
competing subsystem solutions with only limited knowledge
regarding aircraft geometry and other design characteristics.
The methodology presented in this work allows such tradeoffs to
be performed and is driven by subsystem requirements
definition, component modeling and simulation, identification of
critical or constraining flight conditions, and evaluation
of competing architectures at the vehicle and mission level. The
methodology is applied to the flight control actuation
system, where electric control surface actuators are likely to
replace conventional centralized hydraulics in future More
Electric Aircraft. While the potential benefits of electric
actuation are generally accepted, there is considerable debate
regarding the most suitable electric actuator electrohydrostatic
or electromechanical. These two actuator types form
the basis of the competing solutions analyzed in this work,
which focuses on a small narrowbody aircraft such as the
Boeing 737-800. The competing architectures are compared at both
the vehicle and mission levels, using as metrics
subsystem weight and mission fuel burn, respectively. As shown
in this work, the use of this methodology aids the
decision-making process by allowing the designer to rapidly
evaluate the significance of any performance advantage
between the competing solutions.
Keywords
More Electric Aircraft, More Electric Initiative, electric
actuator, electrohydrostatic actuator, electromechanical
actuator
Date received: 28 February 2014; accepted: 13 June 2014
Introduction: Consideration of MoreElectric subsystems in
conceptual design
The conceptual design phase of the traditional aircraftdesign
process typically gives only limited consider-ation to the aircraft
subsystems (In this work, bysubsystems the authors mean Aircraft
EquipmentSystems such as actuation systems for flight
controlsurfaces, landing gear, thrust reversers, nosewheelsteering,
and wheel braking, the environmental con-trol system (ECS) and ice
protection system (IPS),engine starting, etc.), which are addressed
mainly inthe late preliminary and detailed design phases.1
Thisapproach has proven adequate when applied to air-craft with
conventional subsystem solutions due to thepresence of a
continuously updated historical data-base and a relative lack of
coupling/interactionsbetween different subsystems.
This will not be the case, however, for conceptualdesign of More
Electric Aircraft (MEA) or, in thelimit, All Electric Aircraft
(AEA), for which there isno equivalent historical data, and where
subsysteminteractions may be significant. As indicated byFaleiro,
MEA/AEA design will necessitate a newdesign paradigm that considers
the aircraft and itssubsystems as a system of systems.2 More
thoroughconsideration must therefore be given to the
Proc IMechE Part G:
J Aerospace Engineering
2015, Vol. 229(6) 10881102
! IMechE 2014
Reprints and permissions:
sagepub.co.uk/journalsPermissions.nav
DOI: 10.1177/0954410014544303
uk.sagepub.com/jaero
1Aerospace Systems Design Laboratory, School of Aerospace
Engineering, Georgia Institute of Technology, Atlanta, GA,
USA2PACE America Inc., Seattle, WA, USA
Corresponding author:
Imon Chakraborty, Aerospace Systems Design Laboratory, School
of
Aerospace Engineering, Georgia Institute of Technology,
Atlanta,
GA 30332, USA.
Email: [email protected]
at EMBRY RIDDLE AERO UNIV on July 7,
2015pig.sagepub.comDownloaded from
http://pig.sagepub.com/
-
requirements, interactions, and effects of the
aircraftsubsystems even in the conceptual design phase.
This paper presents a requirements-driven method-ology aimed at
MEA/AEA design that integrates theanalysis of subsystem
requirements and effects with theconceptual phase aircraft sizing
process (Figure 1(a)).For each subsystem, the functional
requirements areset by identifying constraining operating
scenarios.The subsystem is sized to those requirements
usingconceptual phase aircraft parameters from the aircraftsizing
analysis (dimensions, estimated takeoff grossweights, etc.).
Fallouts such as subsystem weight(Wsub), non-propulsive power
requirement (Pnp), anddrag coefficient change (CD, if any) can be
fed back tothe aircraft and propulsion system sizing analyses
forthe subsequent iteration (if Cycle? Yes). The con-vergence of
this iterative process yields a design whereconceptual design
parameters such as aircraft weight,wing area, propulsion
requirements, etc. were com-puted by representing subsystem
requirements andeffects explicitly.
This paper focuses on the application of this meth-odology to
electric flight control actuation systems. Asseen in Figure 1(b),
the steps involved mirror the gen-eral steps of Figure 1(a) and
form the subject of subse-quent sections of the paper. An overview
of flightcontrol actuation systems is described in the next
sec-tion. The Control surface actuation requirementssection derives
the actuation requirements of the con-trol surfaces, which feed
into the actuator models that
are described in the EHA and EMAmodels section.The Electric
flight control actuation system architec-tures section identifies
two configurations of interestand develops their corresponding
architectures. TheVehicle level analysis section shows a
comparisonof the two configurations at the vehicle level, basedon
subsystem weight. The Mission level analysis sec-tion compares the
two configurations at the missionlevel, based on fuel burn.
Finally, conclusions drawnfrom this work and avenues identified by
the authorsfor future work are presented.
Overview of flight control actuationsystems: Hydraulic versus
electric
Most commercial and military aircraft in servicetoday use
centralized hydraulic systems for actuatingflight control surfaces,
landing gear, brakes, thrustreversers, etc. These systems,
utilizing electric orengine-driven pumps to pressurize hydraulic
fluid to2035MPa (30005000 psi), have been maturedthrough decades of
aeronautical experience, withredundancy achieved through the
incorporation ofmultiple physically segregated independent
systems.
However, the inevitable technology saturationreached by
hydraulics coupled with significantimprovements in the power
densities of power elec-tronics and electric drives3,4 have led to
a renewedinterest in the research and development of electric
Figure 1. Integrating subsystem sizing into the traditional
aircraft sizing activity in the conceptual design phase. (a)
Methodology
linking subsystem sizing with traditional conceptual design
phase aircraft sizing, with feed-back information flow to enable
design
cycling in response to subsystem changes and (b) Application to
flight control actuation system with electrically actuated
flight
control surfaces.
EHA: electrohydrostatic actuator; EMA: electromechanical
actuator.
Chakraborty et al. 1089
at EMBRY RIDDLE AERO UNIV on July 7,
2015pig.sagepub.comDownloaded from
http://pig.sagepub.com/
-
actuators for flight control actuation as part of theMore
Electric Initiative. Compared to centralizedhydraulics, electric
actuators have the advantage ofPower on Demand, consuming power
only whenthe control surface moves or overcomes a load.Moreover,
they require only electrical energy input(Power By Wire) and allow
the removal of bulkyhydraulic lines, offering potential aircraft
levelweight savings. An additional advantage stems fromthe fact
that electric actuators are Line ReplaceableUnits, and are much
easier to remove/install thanhydraulic actuators.
The majority of research and development activ-ities related to
electric actuation has been on the mili-tary side, starting as far
back as the 1950s with theBritish V-bombers.4 More recent efforts
have beenundertaken in the United States,5 involving theC-141
Starlifter,6 NASAs F-18 Systems ResearchAircraft,7,8 the AFTI
F-16,9 and most recently theF-35 Joint Strike Fighter.10 On the
civil aviationside, a notable MEA is the Airbus A380, where
elec-tric actuators are used for certain control surfaces.11
This 2H/2E configuration has been claimed to offera 450 kg (1000
lb) weight savings compared to a con-ventional 3H configuration.12
With the exception ofthe Joint Strike Fighter, other flight test
programs andanalyses have focused on only single control
surfacetypes at a time, e.g., flaperon,7,8 aileron,6
spoiler,13,14
or high lift devices.15 In this work, since the entireflight
control actuation system must be represented,all the aircraft
flight control surfaces were consideredsimultaneously.
Electric actuator concepts fall mainly into twocategories,
within which there may be architecturalvariationsthe
electrohydrostatic actuator (EHA)and the electromechanical actuator
(EMA). Eachhas advantages and disadvantages relative to theother,
and there is significant debate (and no consen-sus) within the
community regarding the type moreapplicable to flight control
surface actuation. TheEHA and EMA form the competing technologies
ana-lyzed in this work.
The competing subsystem solutions or architectureswere formed by
associating the competing technolo-gies (EHA or EMA) to the
aircraft control surfaces, amultitude of which are present even on
small narrow-body aircraft. These associations were developedtaking
into consideration flight-criticality of the con-trol surfaces,
control surface redundancy, etc.
The current analysis utilized MATLAB/Simulinkand Pacelab SysArc,
an interactive system architec-ture investigation platform
developed by PACEAerospace Engineering and Information
TechnologyGmbH (Organization website: www.pace.de; PacelabSysArc:
www.pace.de/products/preliminary-design/pacelab-sysarc). It
comprises a Knowledge Designerwhere system components (control
surfaces, actu-ators, etc.) were mathematically modeled, and
anEngineering Workbench, using which the
configuration architectures were developed. For add-itional
details regarding the use of Pacelab SysArc foraircraft
architectural analysis, the interested reader isreferred to
Chakraborty et al.16 The proposed meth-odology can of course be
applied using other analysistools as well.
Control surface actuation requirements
Even for aircraft of similar size, the precise control sur-face
configuration is often influenced by organizationaldesign
philosophy, especially for high-lift devices.17 Todemonstrate
thismethodology, the control surface con-figuration and dimensions
of the Boeing 737-800 air-craft were utilized. Each wing contains
an outboardaileron and four flight spoilers for roll control. Aft
ele-vators and a trimmable horizontal stabilizer (THS)provide pitch
control, and a single rudder controlsyaw. The high lift system
comprises two inboardKrueger flaps, four outboard leading edge
slats, andtwo double-slotted trailing edge flaps (TEFs) on
eachwing. The speedbrake function is performed by thesymmetric
deployment of all flight spoilers in the air,and is supplemented on
the ground by the deploymentof two ground spoiler panels on each
wing.
For hinged control surfaces, the required actuationload can be
characterized by the maximum hingemoment Mh,max. The hinge moments
developed forthe ailerons, elevator, and rudder may be expressedin
standard form by
Mh q Sf cf Ch,0 Ch,M eff Ch,M,
1
where q is the dynamic pressure, M is the Machnumber, eff is the
effective angle of attack seen bythe main surface, the control
surface deflection, andSf and cf are the control surface planform
area andchord. Ch,0, Ch,, and Ch, are hinge moment coeffi-cients
representing respectively the effect of main lift-ing surface
camber, eff, and on the hinge moment.
The hinge moment coefficients were determinedbased on the
geometries of the control surfaces andassociated main lifting
surfaces using the empiricalmethodology of Roskam.18 Hinge moments
predictedusing these coefficients were validated against pub-lished
hinge moments from flight tests of the NASAF-18 Systems Research
Aircraft.7 This comparison,shown in Figure 2, yielded acceptable
agreementbetween predicted and observed hinge moments. Forthe
flight and ground spoilers, the method proposedby Scholz19 was used
to determine hinge momentcoefficients.
For each primary control surface, the flight condi-tion
(represented by q, eff, and ) that yields the max-imum hinge moment
can be obtained from therelevant Federation Aviation Regulations
(FARs),and have been identified by Scholz.19 For the aileron,it is
given by FAR 25.349, and corresponds to the
1090 Proc IMechE Part G: J Aerospace Engineering 229(6)
at EMBRY RIDDLE AERO UNIV on July 7,
2015pig.sagepub.comDownloaded from
http://pig.sagepub.com/
-
maximum hinge moment from among three flightconditions abrupt
and full deflection at design man-euver speed (VA), a deflection
yielding the same rollrate at design cruise speed (VC), and a
deflection yield-ing a third of that roll rate at the design dive
speed(VD). For the elevators, FAR 25.255 specifies therequired
positive and negative recovery load factorsthat must be available
following a runaway failureof the THS. For the rudder, the
applicable FARs areFAR 25.149 (one engine inoperative, OEI) and
FAR25.351 (permissible rudder deflections). FAR 25.351and FAR
25.255 were evaluated over the aircraft alti-tude-speed envelope to
determine the most criticalflight condition yielding the maximum
hingemoment (Figure 3). For the flight spoilers, the max-imum hinge
moment occurs when they are extendedduring an emergency descent
performed at the designdive speed (VD). Since weight-on-wheels is
requiredfor ground spoiler deployment, the maximum ratedtire speed
was used to compute their actuation loadrequirement.
High lift device actuating loads strongly depend onflap
mechanism kinematics, which vary significantlybetween both aircraft
and manufacturers.17 In lieu of adetailed analysis of any one
mechanism, the maximumactuation loads for the Krueger flaps and
slats werederived using force and moment coefficients from
pub-lished wind tunnel analyses.20,21 For the TEFs, therequired
actuation load was set to the ratings of theexisting leadscrew
typeactuators (Boeing 737 linear ball-screw flap actuator, Triumph
Gear Systems, TriumphGroup, Inc.,High-LiftActuationSystemsproduct
data-sheet. Available online:
www.triumphgroup.com/companies/triumph-gear-systems-park-city/Documents,
accessed 25 February 2014.).
The maximum angular rates of ailerons, elevators,rudder, and
flight spoilers were set based on theobserved trend of higher rate
requirements formodern automatic flight control systems.22
Representative rates for the high lift devices wereobtained by
timing videos of these devices deploying.Table 1 summarizes the
derived actuationrequirements.
EHA and EMA models
Linear type actuators (as opposed to rotary actuators)were
considered in this work. The relationship
(a) (b)
Figure 2. Comparison of hinge moments predicted using Roskams
method with flight test data from NASAs F-18 SRA. (a) Abrupt
roll doublet - full stick (left): M 0.72, h 7,620 m (25,000 ft),
q-bar 13.60 kPa (284 lbf/ft2) and (b) Abrupt aileron reversal -
full stick(left): M 0.84, h 7,620 m (25,000 ft), q-bar 18.29 kPa
(382 lbf/ft2).
Figure 3. Aircraft flight envelope with maximum hinge
moment flight condition for ailerons, elevators, rudder, and
flight spoilers.
Chakraborty et al. 1091
at EMBRY RIDDLE AERO UNIV on July 7,
2015pig.sagepub.comDownloaded from
http://pig.sagepub.com/
-
between control surface angular movement and actu-ator ram
linear movement for hinged control surfacesis governed by the
kinematics of the control surface toactuator linkage (Figure 4(a)).
This allows the actu-ators output force FL, ram speed v, and ram
positionx to be computed from the control surfaces hingemoment,
angular rate, and angular position.
The actuator is sized based on the required load-speed envelope,
shown notionally in Figure 4(b). Itsstall load F0, no-load
(maximum) speed vmax, andstroke xmax can be computed from the
maximumcontrol surface hinge moment Mh,max, maximumangular rate
_max, control surface angular rangemax, and the linkage kinematics
(Figure 4(a)).
Electrohydrostatic actuator
In the EHA, a bidirectional motor is directly mated toa
bidirectional, fixed-displacement pump, which actu-ates a hydraulic
cylinder whose piston is connected tothe actuator output ram
(Figure 5(a)). The cylinderpressure difference p required to
achieve an outputforce FL, and a piston velocity _x with
acceleration x isgiven by
p 1Apmr x c _x FL 2
where Ap is the piston cross-sectional area, mr themass of
reciprocating components (piston outputrodbalance rod), and c the
viscous friction
(a) (b)
Figure 4. Control surface linkage kinematics and actuator
load-speed envelope. (a) Actuation of hinged control surface using
linear
actuator and (b) actuator load versus speed envelope.
(a)(b)
Figure 5. Notional schematic of electrohydrostatic and
electromechanical actuators. (a) Electrohydrostatic actuator (EHA)
and
(b) electromechanical actuator (EMA).
Table 1. Summary of flight control surface actuation
requirements for example small narrowbody aircraft.
Control surface Actuating load Rate
No.
per aircraft
Ailerons 4,200 Nm 60/s 2
Elevators 7,600 Nm 60/s 2
Rudder 8,200 Nm 60/s 1
Flight spoilers 4,200 Nm 60/s 8
Ground spoilers 3,800 Nm 40/s 4
Trailing edge flaps 51,000 N 102 mm/s 4
Leading edge slats 6,300 N 60 mm/s 8
Krueger flaps 5,600 Nm 16/s 4
1092 Proc IMechE Part G: J Aerospace Engineering 229(6)
at EMBRY RIDDLE AERO UNIV on July 7,
2015pig.sagepub.comDownloaded from
http://pig.sagepub.com/
-
coefficient. The required piston area was computed byevaluating
equation (2) at the stall load condition andwith maximum pressure
difference, i.e., withFL F0, _x x 0, p pmax. Cylinder compo-nents
were sized considering the loads applicable foreach, e.g., piston
rod diameter based on tension/com-pression and buckling, cylinder
wall thickness basedon tensile and circumferential (Hoop) stresses,
etc.23
The cylinder mass mc was computed by summing thecomputed masses
of the constituent components andthe enclosed hydraulic fluid, and
considering thecylinder configuration (single or tandem).
The required total pump flowQtot is given by the sumof the ideal
flow Qideal and the leakage flow Qlkg, i.e.
Qtot Qideal Qlkg Ap _x clkgp 3
where clkg is the leakage coefficient. The pump shaftspeed !p
and required shaft torque p are given by
!p QtotDpump
, p p Dpump
h4
where Dpump is related to the pump displacement(Pump
displacement is defined as the fluid volumepumped per unit
revolution (mm3/rev or in3/rev)) asDpump pump displacement=2 and h
is thepump efficiency. Since the leakage flow is typicallymuch
smaller than the ideal flow, the maximum flowrate corresponds to
the maximum ram velocity, i.e.,Qmaxtot Apvmax. For the EHA, the
motor and pumpshaft speeds and torques are physically identical,
i.e.,!m !p, m p. The required maximum speed andshaft torque of the
motor are then given by
!maxm ApvmaxDpump
, maxm pmax Dpump
h5
A parametric pump model was used to computethe pump mass mp
based on the pump displacement.This included the mass of the
hydraulic fluid presentwithin the pump, while the mass of the
hydraulic fluidin the lines connecting the pump to the cylinder
wasaccounted for using the mass calibration factor dis-cussed in a
subsequent section.
Electromechanical actuator
In the EMA, a bidirectional motor drives a reductiongearbox (GB)
which in turn drives a ballscrew (BS)that converts the rotational
motion of the GB outputshaft to the linear motion of the actuator
output ram(Figure 5(b)). The gearing ratios of the BS and the
GBwere defined as
Gbs !bsv
, Ggb !m!bs
6
where !bs is the angular speed of the rotating memberof the BS.
The gearing ratios can be used to relate
actuator no-load speed and stall load to maximummotor shaft
speed and shaft torque:
!maxm vmax Gbs Ggb, maxm F0
bs Gbsgb Ggb7
where bs and gb are the BS and GB efficiencies. Forthis work, a
GB efficiency of gb 0:9 was used. TheBS efficiency was computed as
bs tanbs=tanbs bs, where bs tan1 bs is the frictionangle and bs
tan11=Gbsrbs is the screw helixangle. The radius of the screw rbs
was selected as thelarger of the radius computed from buckling
consid-eration and that from consideration of combined axialand
torsional loading.23
For the GB, the number of stages required toattain the overall
GB ratio Gbs was computed asn ceil logGgb=log Gmaxstg
, where Gmaxstg 2Lcc=
dmin 1 is the maximum stage reduction ratio for agiven
center-to-center distance Lcc and minimum per-missible pinion
diameter dmin. The GB mass mgb wascomputed after determining the
gear face width fromstrength considerations,23 and accounting for
thenumber of driver-driven pairs.
Motor and power electronics
A brushless DC (BLDC) motor was considered forboth the EHA and
EMA cases, as this type ofmotor has been successfully flight tested
in electricactuators.8 Using parametric BLDC
torque-speedcharacteristics,24 the required motor output
torque(maxout ) was determined based on the following criteria:
1. The peak output torque given by equations (5)and (7) must be
met with a torque margin appliedto satisfy acceleration
requirements against actu-ator inertia.
2. The actuator force-speed (F v) envelope, whenreduced to a
torque-speed ( !) envelope by theactuator gearing, must be
contained within themotor torque-speed envelope.
3. Themotorsmaximumcontinuous torquemustmeetthe actuators
steady-holding load requirement.
Evaluating the input electrical power over theboundary of the
motor ! envelope (accountingfor losses whose magnitudes depend on
the operatingpoint in the envelope), the maximum electrical
inputpower Pmaxin was determined. Torabzadeh-Tari
25 pro-vides an empirical estimate of the mass of a BLDCmotor
based on maxout and P
maxin as
mem 1034:7
maxout
0:841222 26:7 f1 3750:28
0:98 Pmaxin 4910:11kg
8
Chakraborty et al. 1093
at EMBRY RIDDLE AERO UNIV on July 7,
2015pig.sagepub.comDownloaded from
http://pig.sagepub.com/
-
where f1 is the source fundamental frequency. Thesame author
also provides an estimate for theweight of the power electronics
as
mpe 103400
VACin
0:022290 0:3 fsw 60000:8
0:3 Pin 20001:1kg
9
where fsw is the switching frequency, VACin is the input
AC voltage (for this work, the actuators are poweredfrom 115
volts alternating current (VAC) buses), andPinis obtained by
scaling the motors maximum inputpower Pmaxin by the assumed
efficiency of the power elec-tronics. For actuator designs
comprising two motors,each was assumed to have its own power
electronic unit.
Actuator weight computation
Since the component mass estimation described aboveaccounts for
only the major actuator components, butneglects other components
such as fittings, clamps,etc., simply adding these masses would
likely under-predict total actuator mass. To compensate, an
addi-tive (likely under-predicted) estimate of actuator masswas
first obtained by adding the computed masses of
the main components while factoring in the number ofcomponent
units N: present in one actuator, i.e.
mEHA mc Np:mp Nem:mem Npe:mpemEMA mbs Ngb:mgb Nem:mem
Npe:mpe
10
To determine the magnitude of the under-predic-tion, the masses
predicted using equation (10) werecompared against published masses
of electric actu-ators available from previous analyses or flight
tests.For each such case, the actuator was sized to the samestall
load, no-load rate, and stroke as the publishedactuators.
Additional information regarding the actu-ator layout (such as
electrohydraulic (EH) or electro-mechanical redundancy) was
incorporated wheneveravailable (Abbreviations: 3 three phase, PMM
permanent magnet machine, FDP fixed displace-ment pump, EH
electrohydraulic, TC tandemcylinder, BLDCM brushless DC motor, BS
ball-screw, SRM switched reluctance machine, M/P motor/pump, M/GB
motor/gearbox.). A calibrationfactor of 1.15 (i.e., 15%) applied to
the predictedcylinder, pump, BS, and GB masses produced reason-able
agreement, as shown in Tables 2 and 3 for EHAand EMA,
respectively.
Table 2. Comparison of predicted versus published EHA weight
following calibration.
Description F-18 SRA flaperon Typical tandem EHA
Source Navarro7 Sadeghi and Lyons26
Stall load 59.16 kN (13,300 lbf) 142.34 kN (32,000 lbf)
No load rate 195.6 mm/s (7.7 in/s) 203.2 mm/s (8.0 in/s)
Stroke 114.3 mm (4.5 in)
Layout 1 3 PMM! 1 FDP Dual EH channel, 2 TC/surfacePublished
weight 18.8 kg (41.5 lb) 72.3 kg (159.5 lb)
Predicted weight 19.1 kg (42 lb) 74.4 kg (164 lb)
EH: electrohydraulic; PMM: permanent magnet machine; FDP: fixed
displacement pump; TC: tandem cylinder;
3: three-phase.
Table 3. Comparison of predicted versus published EMA weight
following calibration.
Description F-18 SRA flaperon Transport spoiler C-141
aileron
Source Jensen et al.8 Fronista and Bradbury14 Thompson6
Stall load 58.72 kN (13,200 lbf) 222.4 kN (50,000 lbf) 84.74 kN
(19,050 lbf)
No load rate 170.2 mm/s (6.7 in/s) 177.8 mm/s (7.0 in/s) 118.1
mm/s (4.65 in/s)
Stroke 104.8 mm (4.125 in) 152.4 mm (6.0 in) 137.9 mm (5.43
in)
Layout 2 3BLDCM! 1 BS 1 5 SRM! 1 BS 2 M/GB! 1 BSPublished weight
11.8 kg (26 lb) 17.7 kg (39 lb) 15.9 kg (35 lb)
Predicted weight 11.8 kg (26 lb) 18.1 kg (40 lb) 16.8 kg (37
lb)
BLDCM: brushless DC motor; SRM: switched reluctance machine;
M/GB: motor/gearbox; BS: ballscrew; 3: three-phase; 5:
five-phase.
1094 Proc IMechE Part G: J Aerospace Engineering 229(6)
at EMBRY RIDDLE AERO UNIV on July 7,
2015pig.sagepub.comDownloaded from
http://pig.sagepub.com/
-
Electric flight control actuation systemarchitectures
The EHA and the EMA each have certain advantagesand
disadvantages relative to the other, which is thecause of the EHA
versus EMA debate within theflight control actuation community.
If designed to the same stall load, no-load rate, andstroke, the
EMA will generally be lighter than theEHA. This is seen in Figure
6, which comparesweight-optimal EHA and EMA designs for a rangeof
hinge moments. The weight advantage of the EMAis seen to grow with
the magnitude of the sizing hingemoment. The EMA is also
potentially more energy-efficient as it converts electrical power
directly tomechanical power without an intermediate step
ofhydraulic/fluidic power.
However, with regard to flight control surface actu-ation, a
critical hazard for the EMA is the possibility ofcertain
single-point jam failures.27 Since a mechanicaljam is involved, it
is not possible to use multiple EMAs
per control surface to provide redundancy either. Thishas caused
the EMAs suitability for actuating at leastthe flight-critical
control surfaces to be questioned bysome, since a catastrophic
outcome may result fromthese control surfaces jamming at an adverse
position(e.g., fully deflected or hardover position).
Considering this same criterion, the EHA enjoys amajor advantage
in that it can be designed to be fail-safe. If a failure occurs,
the hydraulic cylinder can bedesigned to enter a damped/bypass
mode, preventingthe actuators output ram from jamming but
simul-taneously providing damping against flutter. Thismeans that
EHAs can be used in parallel to meetthe failure probability
criteria applicable to flight-cri-tical control surfaces. They can
also be integrated withcentralized hydraulics (e.g., Electrical
BackupHydraulic Actuator on Airbus A38011).
The EMA may, however, be suitable for the actu-ation of flight
and ground spoilers. The ground spoi-lers are disabled in flight
and are not flight-critical(i.e., their operability is not
necessary for the contin-ued safe flight of the aircraft). Further,
there may becases where the layout and sizes of wing control
sur-faces are such that Failure Mode and Effect Analysis(FMEA) can
be used to establish the fact that theflight spoilers are also not
flight-critical. In suchcases, less stringent failure probability
requirementswould be applicable to these control surfaces, andthey
may presumably be EMA-driven. In fact, EMAdesign for spoiler
actuation was analyzed by Atallahet al.13 and Fronista and
Bradbury.14
A similar argument holds for the high lift devices,and an
EMA-driven high lift system was analyzedexperimentally and from a
safety/reliability stand-point by Bennett et al.15,28 In line with
this reasoning,the two control surface actuation
configurationsshown in Figure 7 were selected for further
analysis.
. Configuration 1: Other than the TEFs, all controlsurfaces are
driven by EHAs. Two parallel EHAs
Figure 7. Configuration 1: TE flaps actuated by EMA-LS. All
other surfaces actuated by EHA. Configuration 2: Ailerons,
elevators,
rudder actuated by EHA. LE slats, Krueger flaps, and spoilers
actuated by EMA. TE flaps actuated by EMA-LS.
EHA: electrohydrostatic actuator; EMA: electromechanical
actuator; TEF: trailing edge flap.
Figure 6. Weight-optimal EHA and EMA designs versus sizing
hinge moment.
EHA: electrohydrostatic actuator; EMA: electromechanical
actuator.
Chakraborty et al. 1095
at EMBRY RIDDLE AERO UNIV on July 7,
2015pig.sagepub.comDownloaded from
http://pig.sagepub.com/
-
are used to drive each aileron and elevator, whilethree are used
for the rudder. Slats, Krueger flaps,flight spoilers, and ground
spoilers are driven by asingle EHA. Each TEF is driven by two
parallelElectromechanically Actuated Leadscrews (EMA-LS:
Electromechanically Actuated Leadscrew. Avariant of the EMA in
which the nut is the trans-lating member instead of the rotating
member.This is similar to the leadscrews used in the con-ventional
Boeing 737-800 TEF system.).
. Configuration 2: The slats, Krueger flaps, flightspoilers, and
ground spoilers are each actuatedusing a single EMA. Actuation for
the remainingcontrol surfaces is identical to Configuration 1.
A detailed functional hazard analysis was beyondthe scope of the
current work, but design practicesfrom Boeing29 and Airbus30 were
adopted to deter-mine the nature of redundancy incorporated into
thetwo architectures.
1. The stringent reliability requirements of the flight-critical
primary control surfaces was met using acombination of actuator
redundancy and internalredundancy.31 Actuator redundancy was
providedby using two parallel actuators in
activeactiveconfiguration for the elevators and ailerons, andthree
parallel actuators in activeactive-standbyconfiguration for the
rudder. Only EHAs wereconsidered for these surfaces due to their
fail-safe characteristics, and parallel EHAs werepowered from
different electric buses. Each EHAdriving a flight-critical control
surface wasassumed to have internal redundancy, in whichtwo
motor-pump pairs drove a tandem cylinder(2M/P ! 1 TC).
2. For non-flight-critical control surfaces such asspoilers,
slats, and Krueger flaps, surface redun-dancy already existed since
there were multiplepanels per wing. In this case, each panel was
pow-ered by a single actuator (either EHA or EMA).The actuators
were connected to the electric busesin a way that ensured symmetric
availabilityof control surfaces following the failure of anyone
bus.
3. TEF panels were provided with actuator redun-dancy, and were
driven by two EMA-LS units atthe two extremities (in practice, this
also preventspanel warping due to aerodynamic loads), pow-ered from
different buses. Each EMA-LS hadinternal redundancy,6 in which two
motor-gearboxpairs drove a BS through a summing element.
Thisarrangement was selected as being representativeof the
Distributed Flap Drive Actuation phil-osophy that is currently
being investigated fortrailing edge high-lift systems in the
interest ofweight savings and functional flexibility.32 Insuch an
arrangement, flap synchronization
between left and right wings will be ensuredthrough electronic
sensing and signalling, in lieuof the mechanically ensured (or
sometimeshydraulically ensured) synchronization found oncommercial
aircraft currently in service.
Vehicle level analysis
For the two configurations shown in Figure 7, a gra-dient-based
constrained optimization was performedto obtain the weight-optimal
EHA and EMA designsfor the relevant control surfaces. For the EHA,
max-imum pump pressure and pump displacement wereused as design
variables, while for the EMA, the BSand GB gearing ratios were
used. For the EHA, thereis an iterative loop between available
pressure differ-ence, cylinder sizing (and mass), and required
pressuredifference. A constraint was therefore enforced toensure
that the required pressure difference did notexceed the pumps
maximum pressure difference.For the EMA, the BS RPM was limited to
a max-imum permissible value.14 For both EHA andEMA, maximum limits
were enforced for motorRPM and permissible current under stall
load.
Table 4 shows the weight breakdown of the twoarchitectures,
which includes the weight of allweight-optimal actuators and also
the associatedwiring from the main electrical buses to the
actuators.The wiring weight was computed by physically pos-itioning
the actuators in Pacelab SysArc, and theninvoking the tools routing
algorithm. This algorithmconverts logical connections between
components(e.g., between a bus and an actuator) to
physicalequivalents, using connecting elements (electricwiring).
Wire gauge is determined from voltage dropand current
considerations, and length is determinedbased on user-defined
allowable pathways, throughwhich electrical wiring is constrained
to run.
Configuration 2 enjoys a weight advantage of5.87% over
Configuration 1, but it should be notedthat this weight advantage
would be only 0.11% ofthe maximum takeoff weight of the Boeing
737-800(79,000 kg).
Mission level analysis
Both configurations were evaluated over the course ofthe
representative mission shown in Figure 8. Primarycontrol surface
movements depend on the phase offlight, and during the mission
simulation, these werecharacterized using the normalized magnitudes
ofdeflection in one or both directions and the cyclesper second
parameter which represents the frequencyof movement.33 The full
control surface deflectionsshown in Table 5 for the Ground segment
representthe control surface check that is performed after
start-ing the engines. The deflections and rates shown forthe
Cruise segment are representative of an
1096 Proc IMechE Part G: J Aerospace Engineering 229(6)
at EMBRY RIDDLE AERO UNIV on July 7,
2015pig.sagepub.comDownloaded from
http://pig.sagepub.com/
-
assumed encounter with moderate turbulence thatlasted for a
fraction of the cruise segment, andrequired more aggressive control
surface movements.Primary control surface movements for the
remainderof the cruise segment were assumed to be negligible.The
high lift devices were assumed to deploy or retractin a manner
consistent with what is observed in atypical commercial aircraft
flight.
To compute the power consumed at the bus by theelectric
actuator, the control surface hinge momentand angular rate were
converted to the requiredforce (Freq) and speed (vreq) at the
control/actuatorlinkage using the control surface gearing Gcs.
Thesewere then propagated through the mathematicalmodel of the
actuator to get the required torque(req) and shaft speed (!m,req)
of the electric motor,effectively using the overall actuator
gearing (Gact).The motor torque and shaft speed requirementswere
then evaluated in conjunction with the assumedtorque-speed
characteristics of the BLDC motor24 tofind the motor input power
requirement. Finally, thiswas scaled using the assumed power
electronics
efficiency pe to give the power draw at the bus (Analternate
method is to estimate the power requirementat the control surface,
i.e., the product of hingemoment and angular rate, and then scale
successivelyby the efficiencies of all actuator components
encoun-tered to obtain the power draw at the bus. However,this
method will predict zero power draw from the busif the control
surface is held fixed against a non-zeroload, while in reality
electric actuators are known toconsume power in this condition. The
methoddescribed above and used in this analysis does notsuffer from
this discrepancy.). This is shown in equa-tion (11) for a
single-channel actuator
Mh,! !Gcs
Freq, vreq !Gact
req,
!m,req !em,!
Pemin !pe
Ppein Pbus
11
Certain control surface motions (e.g., flap
deploy-ment/retraction) occur only during a fraction of theduration
of a mission segment. For these, the averagepower consumption over
the duration of the missionsegment Pseg was computed using the mean
powerconsumed during the motion P and the activity ratio as
follows
tactivetseg
2 0, 1, ) Pseg P 0:1 P
12
As only the flight control actuation subsystem wasbeing
considered, the non-propulsive power require-ments of the other
subsystems (Table 6) were assumedto be identical for both
configurations. The ECS wasassumed to have a bleedless
architecture, and enginebleed air was used only for the cowl IPS,
similar tothe Boeing 787 arrangement. The power and bleedair
demands of these systems were reported by
Table 4. Weight comparison between Configuration 1 and
Configuration 2.
Control surface Actuators Configuration 1 weight (kg)
Configuration 2 weight (kg)
Name No. Act./surf. Design Unit Group Design Unit Group
Aileron 2 2 EHA 15.8 63.2 EHA 15.8 63.2
Elevator 2 2 EHA 25.6 102.4 EHA 25.6 102.4
Rudder 1 3 EHA 27.0 81.0 EHA 27.0 81.0
Flight spoiler 8 1 EHA 10.2 81.6 EMA 9.0 72.0
Ground spoiler 4 1 EHA 8.4 33.6 EMA 8.9 35.6
Krueger flap 4 1 EHA 19.9 79.6 EMA 13.4 53.6
LE slat 8 1 EHA 16.3 130.4 EMA 9.7 77.6
TE flap 4 2 EMA-LS 44.5 356.0 EMA-LS 44.5 356.0
Total weight of actuators! Total! 927.8 Total! 841.4Electrical
wiring Length (m) AWG Weight (kg) Length (m) AWG Weight (kg)
571 4/0 544 571 4/0 544
Total architecture weight! Total! 1471.8 Total! 1385.4
EHA: electrohydrostatic actuator; EMA-LS: electromechanically
actuated leadscrew; EMA: electromechanical actuator.
Figure 8. Representative flight profile for mission level
ana-
lysis. In addition to the 3000 NM (380 min) flight shown,
shorter flights of 350 NM (63 min), 850 NM (132 min), and
1350 NM (202 min) were also evaluated with representative
cruise altitudes.
Chakraborty et al. 1097
at EMBRY RIDDLE AERO UNIV on July 7,
2015pig.sagepub.comDownloaded from
http://pig.sagepub.com/
-
Tagge et al.34 for a Boeing 767-type airplane, andwere scaled
using guidelines reported by Cleveland35
to Boeing 737-800 size. Power demands for instru-mentation,
communication, navigation (tabulated asinstruments), equipment and
furnishings, lights,pumping fuel, and miscellaneous loads were also
setin a similar manner. The power demand for theassumed electrical
wing de-icing system was estimatedusing a rapid evaluation method
developed by Meierand Scholz36 which was shown to have good
agree-ment with the Boeing 787 wing IPS power demand.Centralized
hydraulics were assumed to be retained toactuate the landing gear,
landing gear bay doors,brakes, nosewheel steering, and cargo doors.
Thepower consumption for these users for differentflight phases was
reported by de Tenorio37 for thecase of the similar-sized Airbus
A320, and has beenincorporated as a shaft-power offtake (to run
engine-driven hydraulic pumps) in this work.
The propulsion system characteristics were gener-ated using the
Environmental Design Space tool38
developed for the U.S. Federal AviationAdministrations Office of
Environment and Energy(FAA/AEE) and NASA. At the core of this tool
is the
Numerical Propulsion System Simulation (NPSS)39
tool, which was used to generate an engine deck thatis
representative of a CFM56-7B27 type engine, butdoes not utilize or
contain any proprietary information.
The difference between the two configurations interms of mission
fuel burn may be caused by a differ-ence in either the
non-propulsive power requirementor the subsystem weight. While both
effects were con-sidered in this work, the magnitude of the first
wasquite small, as seen from Table 7. Further, the peakpower demand
for flight control actuation (a factorinfluencing generator sizing)
does not occur for anyof the flight conditions of Table 7. Tagge et
al.34
note that this critical peak power flight condition isgenerally
the OEI crosswind landing condition,where the combination of high
load and rate require-ments justifies a conservative sizing
approach based oncorner power (Corner power for a control surface
isthe product of maximum hinge moment and max-imum angular rate.).
Referring to Table 1 and assum-ing operability of both ailerons,
both elevators, therudder, and four out of eight flight spoilers in
theOEI condition, the combined corner power evaluatesto around 51
kW for both configurations.
Table 6. Non-propulsive power requirements (excluding flight
control actuation).
System Type Unit Ground Takeoff Climb Cruise Descent Landing
OEI
ECS Bleedless kVA 129 129 210 210 210 129 204
Instruments Electric kVA 8 8 8 8 8 8 8
Equipment Electric kVA 60 60 60 65 0 0 0
Fuel EMDP kVA 7.5 7.5 7.5 7.5 7.5 7.5 7.5
Fuel EDP kW 7.5 7.5 7.5 7.5 7.5 7.5 7.5
Wing IPS Electric kVA 17 17 25 25 25 17 17
Cowl IPS Bleed air kg/s 2.2 2.2 2.2 0 2.2 2.2 1.1
Lights Electric kVA 10 10 10 10 10 10 10
Hyd. users EDP kW 1.43 0 0.67 0 0.44 9.83 9.83
Misc. Electric kVA 5 5 5 5 5 5 5
Demand Electric kVA 124 124 171 173 139 92 260
per Mechanical kW 4.9 4.1 4.5 4.1 4.4 9.5 15.2
Engine Bleed kg/s 1.1 1.1 0 0 1.1 1.1 1.1
ECS: environmental control system; OEI: one engine inoperative;
IPS: ice protection system; EMDP: electric motor driven pump; EDP:
engine driven
pump.
Table 5. Characterization of aileron, elevator, rudder, and
flight spoiler movements.
Control! Aileron Elevator Rudder Flt. spoiler
Motion! / cps / cps / cps cpsGround 1/1 0.4 1/1 0.4 1/1 0.4 1
0.4Takeoff 0.12/0.2 0.4 0.2/0.2 0.4 0.2/0.2 0.4 0.12 0.4Climb
0.12/0.2 0.2 0.2/0.2 0.2 0.2/0.2 0.2 0.12 0.2Cruise 0.5/0.5 0.4
0.3/0.3 0.4 0.3/0.3 0.4 0.5 0.4Descent 0.12/0.2 0.3 0.2/0.2 0.3
0.2/0.2 0.3 0.12 0.3Landing 0.4/0.53 0.3 0.53/0.53 0.3 0.53/0.53
0.3 0.4 0.3
1098 Proc IMechE Part G: J Aerospace Engineering 229(6)
at EMBRY RIDDLE AERO UNIV on July 7,
2015pig.sagepub.comDownloaded from
http://pig.sagepub.com/
-
The difference in mission fuel burn is primarilydriven by the
difference in weight between the twoconfigurations (Table 4). In
addition to the3000NM mission shown in Figure 8, several
inter-mediate trip distances that are more commonlyencountered in
small narrowbody aircraft operationswere also considered, for which
cruise altitudes wereselected based on a review of typical cruise
altitudesused by airlines for such distances (Online live
flighttracking resource, www.flightaware.com). The com-parison of
mission fuel burn for all flight profiles con-sidered is presented
in Table 8.
Conclusions and future work
Motivated by the need to consider MEA or AEA sub-systems in
greater detail in the conceptual designphase, a methodology was
presented that integratesthe analysis of subsystem requirements and
effectswith the conceptual stage aircraft sizing process.
This requirements-driven methodology analyzesand sizes the
subsystems based on functional require-ments in conjunction with
aircraft design parametersthat are either available or easily
estimable in the con-ceptual design phase. Since information
regardingsubsystem characteristics can be fed back into the
air-craft and propulsion system sizing process, the con-verged
design point obtained from this iterativescheme is one where the
effect of aircraft characteris-tics on the subsystems and the
feedback effect of sub-system characteristics on aircraft and
propulsor sizinghave been accounted for explicitly. Such a
method-ology will give the conceptual phase designer ofMEA/AEA the
ability to rapidly evaluate the aircraftand mission level effects
of subsystem architectures,
allowing the subsystem concept space to be exploredmore
thoroughly. The designer will therefore be ableto make more
informed decisions regarding subsys-tem architectures before
progressing to the subse-quent design phases, where design freedom
isprogressively reduced as configuration characteristicsare frozen,
and after which major design changesbecome expensive in terms of
both time and cost.
The methodology was applied to the conceptualdesign phase
analysis of the flight control actuationsystem of a small
narrowbody aircraft (e.g., Boeing737 or Airbus A320). Rather than
comparing theeffect of transitioning from hydraulic to
electricflight control actuation systems, this paper comparedtwo
candidate electric actuation solutions. It wasfound that a
configuration that employed EHAs forflight-critical control
surfaces and EMAs for non-flight-critical ones enjoyed a marginal
aircraft-levelweight advantage over another
predominantly-EHAconfiguration. A concomitant fuel burn
advantage,driven primarily by the weight advantage, was
alsodetected. The conceptual phase designer exercisingthis
methodology would be in a position to weighthe observed advantages
against additional consider-ations such as safety/reliability. In
this specific case,considering the possibility of a single-point
jammingfailure of an EMA and the potentially hazardousflight
conditions that may result from a jammed spoi-ler or high-lift
device, the designer may be inclined toaccept the marginal weight
penalty of the predomi-nantly-EHA configuration in exchange for its
fail-safe nature.
Additional factors that were not directly con-sidered in this
work would also significantly influencesuch a design decision.
Given that electric motors ingeneral and power electronics in
particular are quitesensitive to operating temperature from the
point ofview of performance, reliability, and life, thermal
per-formance requirements would be significant designdrivers for
electric actuators. The thermal challengeis compounded by the
absence of centralized hydrau-lic fluid to act as a heat sink, the
likely infeasibility ofactive cooling systems for actuators from a
weight
Table 7. Actuation power requirements for PFCS and SFCS for
Configurations 1 and 2 (averaged
over mission segment duration).
SegmentConfiguration 1 (kW) Configuration 2 (kW)
PFCS SFCS Total PFCS SFCS Total
Ground 0.02 1.19 1.21 0.02 1.20 1.22
Takeoff 0.24 0.22 0.46 0.23 0.25 0.48
Climb 0.67 0.55 1.22 0.67 0.59 1.26
Cruise 0.99 0.99 0.98 0.98
Turbulence 4.96 4.96 4.92 4.92
Descent 0.97 0.54 1.51 0.96 0.57 1.53
Landing 3.26 0.22 3.48 3.24 0.26 3.50
PFCS: primary flight control surfaces; SFCS: secondary flight
control surfaces
Table 8. Mission fuel-burn comparison (Configuration 2
relative to Configuration 1).
Trip Distance! 350 NM 850 NM 1350 NM 3000 NM
Block fuel (%) 0.06 0.08 0.09 0.11
Chakraborty et al. 1099
at EMBRY RIDDLE AERO UNIV on July 7,
2015pig.sagepub.comDownloaded from
http://pig.sagepub.com/
-
perspective, and the degradation in the convectiveheat transfer
coefficient with altitude. In order to befeasible therefore, an
electric actuator must bedesigned to tolerate the most constraining
thermalloads (which can be initially misidentified8) with pas-sive
cooling alone. Clearly, a thermal performanceadvantage enjoyed by
an actuator would significantlyinfluence a design decision.
Based on the experience and results obtained fromthis work, the
authors have identified three main ave-nues for further research,
which will be the subjects offuture publications. The first avenue
of research will beaimed at determining how the relative
performance ofthe two actuation configurations varies as
aircraftgross weight increases. In other words, how
significantwould the weight difference be if, instead of a
Boeing737 or Airbus A320-sized aircraft, a Boeing 747-sizedor
Airbus A380-sized aircraft is considered? As seenfrom Figure 6, the
magnitude of the EMAs weightadvantage over the EHA increases as the
actuationload increases. However, not only do larger-sized
air-craft have larger overall control surface planformareas, they
also generally have multiple control surfacesections (e.g., a split
rudder, or multiple aileron sec-tions per wing). The actuators
driving these split sur-faces would be sized to smaller loads, but
they wouldbe larger in number. Second, some aircraft may
havecontrol surfaces such as inboard high-speed ailerons(e.g.,
Boeing 747) that others do not (e.g., Boeing 737).Thus instead of
attempting to scale the results (whichwould disregard or mask these
configurational differ-ences), future work will attempt to bound
the prob-lem by assessing the two configurations applied to asmall
number of aircraft which taken together encom-pass a broad range of
aircraft gross weights.
The second avenue of research will consider otheraircraft
subsystems such as the ECS, IPS, etc. in add-ition to the flight
control actuation system, using thesame requirements-driven
subsystem sizing methodsdiscussed in this work. This will allow the
conceptualdesigner to set mission requirements (payload,
range,etc.) and compare a spectrum of aircraft
subsystemarchitectures. For example, at the two extremesmight be an
AEA and another candidate with conven-tional subsystem
architecture, with several MEAoptions in between. More Electric
subsystem solutionsapplied to multiple subsystems may in this case
resultin large differences in the non-propulsive powerrequirements
and weights of subsystems relative to abaseline with conventional
architecture. In this case,the feedback loop of the proposed
methodologywill allow the snowballing effects of the
subsystemchanges on the overall aircraft characteristics to
becaptured. Depending on the case at hand, the differ-ence in
aircraft characteristics between uncycled(feedback loop not active)
and cycled (feedbackloop active) designs may be quite substantial,
andsomething of considerable interest to the conceptualphase
designer.
The third avenue of research will consider the effectof
uncertainty in the decision-making process, inother words, it will
deal with decision making in thepresence of uncertainty. This is
predicated on the factthat in the conceptual design phase there are
signifi-cant levels of uncertainty associated with the param-eters
and metrics that the current analysis is based on.For example, the
actuator sizing process is driven bythe estimation of control
surface aerodynamic loads.These loads are significantly affected by
design par-ameters that are seldom considered at the
conceptualdesign phase (e.g., flap mechanism kinematics,
etc.).Therefore, rather than having a single sizing load, onewould
in reality have a probabilistic distribution ofloads that would
feed into the actuator sizing process.This uncertainty, coupled
with uncertainties in com-ponent material properties, etc., would
ultimatelyyield distributions for the metrics of interest
(weight,fuel-burn, etc.) on which the design decisions will
bebased. The explicit consideration of these uncertain-ties in the
decision-making process will allow for adecision that is more
robust, i.e., less sensitive to theeffect of the uncertainties.
Funding
This research received no specific grant from any fundingagency
in the public, commercial, or not-for-profit sectors.
Conflict of interest
None declared.
Acknowledgements
The authors would like to thank Mr. Christopher Perullo,Research
Engineer II at the Aerospace Systems Design
Laboratory, for supporting their investigation on the
pro-pulsion side with his expertise and experience with the EDSand
NPSS tools.
References
1. Raymer DP. Aircraft design: a conceptual approach. 4thed.
Reston, Virginia: American Institute of Aeronautics&
Astronautics, 2002.
2. Faleiro L. Beyond the more electric aircraft. Aerosp Am2005;
43: 3540.
3. Cronin MJJ. The all-electric aircraft. IEE Rev 1990;
36:309311.
4. Jones R. The more electric aircraft assessing the bene-fits.
Proc IMechE, Part G: J Aerospace Engineering 2002;216: 259269.
5. Cloyd JS. Status of the United States air forces moreelectric
aircraft initiative. Aerosp Electron Syst Mag,IEEE 1998; 13:
1722.
6. Thompson K. Notes on The electric control of largeaeroplanes.
Aerosp Electron Syst Mag, IEEE 1988; 3:1924.
7. Navarro R. Performance of an Electro-HydrostaticActuator on
the F-18 Systems Research Aircraft. 16thDigital Avionics Systems
Conference, Irvine,California, USA, October 1997.
1100 Proc IMechE Part G: J Aerospace Engineering 229(6)
at EMBRY RIDDLE AERO UNIV on July 7,
2015pig.sagepub.comDownloaded from
http://pig.sagepub.com/
-
8. Jensen SC, Jenney GD and Dawson D. Flight testexperience with
an electromechanical actuator on theF-18 systems research aircraft.
In: Proceedings of the
19th digital avionics systems conference, 2000, vol. 1,pp.2E31.
Philadelphia, Pennsylvania, USA: IEEE.
9. Burkhard A and Dietrich R. Joint strike fighter
integrated subsystems technology, a demonstra-tion for industry,
by industry. J Aircraft 2003; 40:906913.
10. Vohnout S, Bodden D, Kim BU, et al. Prognostic-enabling of
an electrohydrostatic actuator (EHA)system. In: Annual conference
of prognostics andhealth management society,
http://72.27.231.73/sites/
phmsociety.org/files/phm_submission/2012/phmc_12_138.pdf (2012,
accessed 9 July 2014).
11. Van Den Bossche D. The A380 flight control electrohy-
drostatic actuators, achievements and lessons learnt.
In:Proceedings of the 25th congress of the international coun-cil
of the aeronautical sciences (ICAS), 38 September
2006, Hamburg, Germany, vol. 7, p.1. Paper ICAS2006-7.4.1.
http://www.icas.org/ICAS_ARCHIVE/ICAS2006/ABSTRACTS/048.HTM.
12. Kulshreshtha A and Charrier J. Electric actuation forflight
and engine control: evolution and challenges. In:SAE-ACGSC meeting
99,
http://acgsc.org/Meetings/Meeting_99/SubcommitteD/Electric_Actuation-
SAE%20ACGSC-Boulder%20Mtg%202007.pdf (2007,accessed 9 July
2014).
13. Atallah K, Caparrelli F, Bingham CM, et al. Permanent
magnet brushless drives for aircraft flight control sur-face
actuation. In: IEE colloquium on electrical machinesand systems for
the more electric aircraft (Ref. No. 1999/
180), 1999, pp.8/18/5. London, UK.
http://ieeexplore.ieee.org/xpl/articleDetails.jsp?arnumber=828628.
14. Fronista GL and Bradbury G. An electromechanicalactuator for
a transport aircraft spoiler surface. In:
Proceedings of the 32nd intersociety energy
conversionengineering conference, IECEC-97, 1997, vol. 1,pp.694698.
Honolulu, HI, USA. http://ieeexplore.
ieee.org/xpls/abs_all.jsp?arnumber=659274.15. Bennett JW, Mecrow
BC, Jack AG, et al. A prototype
electrical actuator for aircraft flaps. IEEE Trans Indus
Appl 2010; 46: 915921.16. Chakraborty I, Mavris D, Emeneth M, et
al. A system
and mission level analysis of electrically actuated flight
control surfaces using Pacelab SysArc. In: AIAA scienceand
technology forum and exposition (SciTech 2014),AIAA-2014-0381,
http://arc.aiaa.org/doi/abs/10.2514/6.2014-0381 (2014, accessed 9
July 2014).
17. Rudolph PKC. High lift systems on subsonic
commercialairliners. NASA Contractor Report 4746, 1996. NASAAmes
Research Center; Moffett Field, CA, United
States: National Aeronautics and Space
Administration.http://ntrs.nasa.gov/search.jsp?R=19960052267.
18. Roskam J. Airplane design part VI preliminary cal-
culation of aerodynamic, thrust and power characteris-tics. Des
Anal Res 1999; 5: 463490.
19. Scholz D. Development of a CAE-Tool for the designof flight
control and hydraulic systems. In: Institution
of Mechanical Engineers (ed.) Avionic Systems, Designand
Software. London: Mechanical EngineeringPublications, 1996,
pp.122.
20. Moraris VM, Lawson NJ and Garry KP. Aerodynamicand
performance characteristics of a passive leading
edge Krueger flap at low Reynolds numbers. AeronautJ 2012; 116:
759769.
21. Kelly JA and McCullough GB. Aerodynamic loads on a
leading-edge flap and a leading-edge slat on the NACA64A010
airfoil section. NACA TN 3220, 1954.California, USA: National
Advisory Committee for
Aeronautics (NACA).
http://naca.central.cranfield.ac.uk/reports/1954/naca-tn-3220.pdf.
22. Charrier JJ and Kulshreshtha A. Electric actuation for
flight & engine control system: evolution, current
trends& future challenges. In: Collection of technical
papers-45th AIAA aerospace sciences meeting, AIAA-2007-1391, 2007,
pp.1642116440. Reno, Nevada, USA.
http://arc.aiaa.org/doi/book/10.2514/MASM07.23. Khurmi RS and
Gupta JK. A textbook of machine
design. Ram Nagar, New Delhi, India: Eurasia
Publishing House (Pvt.) Ltd., 2005.24. Frischemeier S.
Electrohydrostatic actuators for air-
craft primary flight control types, modelling and
evaluation. In: Proceedings of the 5th Scandinavianinternational
conference on fluid power (SICFP 1997),Linkoping, Sweden, 2830 May
1997, pp.2830.
http://doku.b.tu-harburg.de/volltexte/2006/301/.25.
Torabzadeh-Tari M. Dimensioning tools of MEA actu-
ator systems, including modeling, analysis and
technologycomparison. Doctoral Dissertation, KTH Electrical
Engineering, 2008.26. Sadeghi T and Lyons A. Fault tolerant EHA
architec-
tures. Aerosp Electron Syst Mag, IEEE 1992; 7: 3242.
27. Botten SL, Whitley CR and King AD. Flight controlactuation
technology for next-generation all-electric air-craft. Technol Rev
J 2000; 8: 5568.
28. Bennett JW, Mecrow BC, Atkinson DJ, et al. Safety-critical
design of electromechanical actuation systems incommercial
aircraft. Electric Power Appl, IET 2011; 5:3747.
29. Yeh Y. Triple-triple redundant 777 primary flight com-puter.
In: Proceedings of aerospace applications confer-ence, 1996, vol.
1, pp.293307. Aspen, Colorado, USA:
IEEE.30. Traverse P, Lacaze I and Souyris J. Airbus fly-by-wire:
a
process toward total dependability. In: P Traverse, I
Lacaze and J Souyris (eds) Building the information soci-ety.
Airbus, France: Springer, 2004, pp.191212.
http://www.icas.org/ICAS_ARCHIVE/ICAS2006/
ABSTRACTS/050.HTM.31. Bennett J, Mecrow B, Atkinson D, et al.
Failure
mechanisms and design considerations for fault tolerantaerospace
drives. In: 2010 XIX international confer-
ence on electrical machines (ICEM), 2010, pp.16.Rome, Italy:
IEEE.
32. Recksiek M. Advanced high lift system architecture
with distributed electrical flap actuation. In: Aviationsystem
technology workshop,
http://www.transport-research.info/Upload/Documents/201210/20121030_
122633_15453_Advanced_High_Lift_System_Architecture_with_Distributed_Electrical_Flap_Actuation.pdf
(2009, accessed 9 July 2014).
33. Simsic CJ. Electric actuation system duty cycles. In:
Proceedings of the IEEE 1991 national aerospace andelectronics
conference (NAECON 1991), 1991,pp.540545. Dayton, Ohio, USA:
IEEE.
34. Tagge GE, Irish LA and Bailey AR. BoeingCommercial Airplane
Company, Seattle, WA, USA.
Chakraborty et al. 1101
at EMBRY RIDDLE AERO UNIV on July 7,
2015pig.sagepub.comDownloaded from
http://pig.sagepub.com/
-
Prepared for NASA Langley Research Center,Hampton, VA, USA,
under Contract
NASI-17528.http://www.dtic.mil/dtic/tr/fulltext/u2/a304177.pdf.
35. Cleveland FA. Size effects in conventional aircraftdesign. J
Aircraft 1970; 7: 483512.
36. Meier O and Scholz D. A handbook method for the
estimation of power requirements for electrical de-icing
systems. In: DLRK, Hamburg,
http://www.mp.haw-hamburg.de/pers/Scholz/MOZART/
MOZART_PUB_DLRK_10-08-31.pdf (2010, accessed9 July 2014).
37. De Tenorio C. Method for collaborative conceptualdesign of
aircraft power architectures. Doctoral
Dissertation, Georgia Institute of Technology, 2010.
38. Kirby M and Mavris D. The environmental designspace. In:
26th international congress of the aeronauticalsciences including
the 8th AIAA Aviation Technology,
Integration, and Operations (ATIO) Conference, 1419September
2008, Anchorage, Alaska, USA. PaperICAS 2008-4.7.3, pp.1419.
http://www.icas.org/
ICAS_ARCHIVE/ICAS2008/ABSTRACTS/586.HTM.
39. Lytle JK. The Numerical Propulsion System Simulation:
An Overview. Cleveland, OH, USA: NationalAeronautics and Space
Administration GlennResearch Center, 2000,
NASA/TM-2000-209915.http://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/
20000063377.pdf.
1102 Proc IMechE Part G: J Aerospace Engineering 229(6)
at EMBRY RIDDLE AERO UNIV on July 7,
2015pig.sagepub.comDownloaded from
http://pig.sagepub.com/