Engine Design Studies for a
`
`
`Silent Aircraft
`
`
`
`
`Cesare A. Hall
`
`
`Daniel Crichton
`Department 01 Enginuerlng.
`
`
`Whittle Laboratory.
`
`Marlingley Road.
`
`Cambridge. U K
`
`
`
`
`
`
`
`Thc Silerrt Ar't-craft lnr'rt'atr'r'c t'.r tr rt-.rt'rrrt‘lr prrtjectfurrdcd by the Cambridge-MIT Institute
`
`
`
`
`
`
`
`
`
`
`
`
`aimed at rzducfrlg urn-mfr tn.-r‘.r¢' rn rhr pm'm‘ whet-1' it
`is frrrpcrtrptibfc in the urban
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`Ellt'fi’tJ.t1.rt'I€nrS around ut'rports. Tlrc pntpulsthn .r_v.rtcrrr being aft-velopcdfor this project has
`
`
`
`
`
`
`
`
`
`
`
`a rhenntId_r'ntrrnic t'_rr'l'e based on on ulrrafrigh bypass ratio turbofan combined wttlr u
`
`
`
`
`
`
`
`
`
`
`
`
`
`t'r:rr'uhle area e.t'lrau.rt' rt.-r.-,:lc and on rmhcdrlml t'rt.rtullatt'tm. This cycle has been matched
`
`
`
`
`
`
`
`
`
`
`
`
`
`to rhcfligltr rt-tr'.r.rr'orr and rltrrrst mqttr'rt°ntenr.r nfrtn nll»lrfit'ng body at'rj‘i'anre. and thmrtgh
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`precrlre .rr:hedrtl'r'ng of the vrrrirrhlc t'.t'lrurr.rr rm::_t‘r'.
`the origin: operating conditions have
`
`
`
`
`
`
`
`
`
`
`
`
`been oprtmfzedfor marfmtmr .tltrt.=.vr at mp-of-t-lt'nrh. minimum file! consurrtptirrn during
`
`
`
`
`
`
`
`
`
`
`
`('l"l4'lS£’. and rrrtrtirrrttm jet nor‘.rc' or fair nfritrrdc. Tlrr'.r paper pmpo.rc.r engine mechanical
`
`
`
`
`
`
`
`
`
`
`
`
`
`ctrra.-rgements that am ntcer the r'_\-rte rcr,rtrt'rmt:=rrts and. trim: fmrralled in an appropriate
`
`
`
`
`
`
`
`
`
`
`
`
`
`oirjfrornc. will be quiet
`rtclcrrft-c m rt.-rrt-tit rm'lmfarr.r. To rvrdut-e the engine n-eight, a
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`s_v.r'rem with (J gearbox. or sortie orltcrfrtnn of .rl'm_ft‘ .rpr-ed rt-dut-rim: dftllff. is pmposcd.
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`Tlrr'.r is r.‘rmtbEned tr-Etlt
`fI'J'W-.l‘pt’t’t‘.il fun ttml rt rurlalrte will:
`ltfgh gap-chord spacing to
`rt
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`further reduce rurhrrnrachfnerw .rrm.-re rmr'se. Art engine t‘r.~rtfi_r{ttrt:ttr'orr
`three farts
`tr-‘I'll:
`
`
`
`
`
`
`
`
`
`
`
`dritren by a single core is (J'f.\‘rJ _m‘(’.l‘l’fltt?(l. and this is c.tpct'te(l' In lmt'e_ftcrrher weight, fuel
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`brim. and not'.rt= benefits. [DOII Ill] 1 lfifl .247239ll]
`
`
`
`
`
`
`
`Introduction
`
`
`
`
`
`The Silent A'rrr:ml't Initiative is a multidisciplinary project that
`
`
`
`
`
`
`
`
`
`is developing a concept aircraft with noise emission as the pri-
`
`
`
`
`
`
`
`
`
`
`rnary design driver. The aircraft is aimed at entry into service in
`
`
`
`
`
`
`
`
`
`
`
`
`about 20 years. and the ambitious objective is to reduce the noise
`
`
`
`
`
`
`
`
`
`
`
`
`generated to the point where it would be imperceptible above
`
`
`
`
`
`
`
`
`
`
`background noise in El typical urban environment outside an air-
`
`
`
`
`
`
`
`
`
`pon. Such an aircraft could be deemed as “silent." and this would
`
`
`
`
`
`
`
`
`
`
`
`represent :1 reduction in aircraft noise grtratcr than that achieved
`
`
`
`
`
`
`
`
`
`over the last fifty yours. l-Ttgurc I illustrates the scale of this chal-
`
`
`
`
`
`
`
`
`
`
`
`
`lenge. showing thc Silent Aircrtrft noise target relative to the com-
`
`
`
`
`
`
`
`
`
`
`ponent noise levels for it current passenger aircraft. Note that to
`
`
`
`
`
`
`
`
`
`
`
`reach this noise level requires an aircraft that is less than half as
`
`
`
`
`
`
`
`
`
`
`
`
`
`noisy us the target idcntificd by the ACARE vision for 2020
`
`
`
`
`
`
`
`
`
`
`
`
`In order to reach the Silent Aircraft noise goal. large reductions.
`
`
`
`
`
`
`
`
`
`
`
`relative to currtznt technology. are required for all components of
`
`
`
`
`
`
`
`
`
`
`engine and airframe noise. To make such large rcductions. several
`
`
`
`
`
`
`
`
`
`
`methods must be employed simultaneously [see [23]). For ex-
`
`
`
`
`
`
`
`
`ample. to reduce jet noise. It very large. low-velocity exhaust flow
`
`
`
`
`
`
`
`
`
`
`
`is requircd combined with a pt>wcr~managcmt:nt departure proce-
`
`
`
`
`
`
`
`the
`dure. To makt: atlcqutttc turhnmachincry noise reductions.
`
`
`
`
`
`
`
`
`source noise can be reduced with improved component design and
`
`
`
`
`
`
`
`
`
`
`new engine configurations, but further attenuation of the noise is
`
`
`
`
`
`
`
`
`
`
`also needed using acoustic liners and shielding by the airframe
`
`
`
`
`
`
`
`
`
`
`[4].
`
`in addition to the aggressive noise target. the new aircraft must
`
`
`
`
`
`
`
`
`
`
`
`be economical relative to other aircraft of the future. This requires
`
`
`
`
`
`
`
`
`
`
`
`
`it propulsion system that has competitive fuel burn as well as
`
`
`
`
`
`
`
`
`
`
`acceptable development. acquisition. and maintenance costs. Prior
`
`
`
`
`
`
`
`to the work in this paper. several trade studies were completed to
`
`
`
`
`
`
`
`
`
`
`
`
`dctemtjnc the potential noise reductions possible for various en-
`
`
`
`
`
`
`
`
`gine configunrtions and to understand their
`implications for
`
`
`
`
`
`
`
`
`weight and performance [2]. This work found that a propulsion
`
`
`
`
`
`
`
`
`
`
`system embedded into the rear upper surface of an all-lifting body
`
`
`
`
`
`
`
`
`
`
`
`was best suited to meeting the project objectives. Furtltcnrtoro. a
`
`
`
`
`
`
`
`
`
`
`
`
`
`turbofun system with at variable exhaust was shown to have the
`
`
`
`
`
`
`
`
`
`
`
`potential to have lower fuel consumption for a given noise level.
`
`
`
`
`
`
`
`
`
`
`
`Several previous research projects have also studied advanced
`
`
`
`
`
`
`
`
`UHBR engine conligurations aimed at significant improvements
`
`
`
`
`
`
`
`in noise nndlor fuel consumption. For example. the NASA study
`
`
`
`
`
`
`
`
`
`
`of advanced engines for high efticiency [5] looked at several con-
`
`
`
`
`
`
`
`
`
`
`figurations. including geared fans and contm-fart designs. aimed at
`
`
`
`
`
`
`
`
`
`weight and fuel burn reductions. Another system study of engine
`
`
`
`
`
`
`
`
`
`
`concepts carried out by NASA [6] investigated the optimum en-
`
`
`
`
`
`
`
`
`
`gine pzmtrnctcrs for low noise with acceptable operating costs.
`
`
`
`
`
`
`
`
`The dcsigrt considerations for a new UHBR engine. aimed at re-
`
`
`
`
`
`
`
`
`
`
`duced fuel burn. arc clearly outlined in [T]. and Ref. [8] gives a
`
`
`
`
`
`
`
`
`
`
`
`
`
`good overview of fururc technology required to furtlter reduce
`
`
`
`
`
`
`
`
`
`noise from conventional aircraft engines. This proposes the use of
`
`
`
`
`
`
`
`
`
`
`geared turbofans to give a large improvement in noise entjssion.
`
`
`
`
`
`
`
`
`
`A study of more radical propulsion concepts for a fimctionally
`
`
`
`
`
`
`
`
`
`
`silent aircraft is also presented in [9}. which proposes distributed
`
`
`
`
`
`
`
`
`
`
`engine systems integrated with at blended-wing-body type aircraft.
`
`
`
`
`
`
`
`
`What is new in the present study is that the off-design perfor-
`
`
`
`
`
`
`
`
`
`
`
`mance ol' the engine has been considered from the start of the
`
`
`
`
`
`
`
`
`
`
`
`
`design process. This is key since the engine conditions when low
`
`
`
`
`
`
`
`
`
`
`
`noise is essential are fat‘ from the design point (typically top-of-
`
`
`
`
`
`
`
`
`
`
`
`climb or cruise}. In addition. the engine cycle in this project has
`
`
`
`
`
`
`
`
`
`
`
`
`been optimized for operation with a variable exhaust system and
`
`
`
`
`
`
`
`
`
`
`for an installation embedded within rm all-lifting wing typo air-
`
`
`
`
`
`
`
`
`
`frame. Prcvious studies have tended to focus on engine designs
`
`
`
`
`
`
`
`
`
`
`intended for conventional tube-und—wing aircraft.
`
`
`
`
`
`The current paper. lltercfttrc. aims to extend the previous work
`
`
`
`
`
`
`
`
`
`
`[2]. which was based on quite simple analyses.
`to create more
`
`
`
`
`
`
`
`
`
`
`
`dctailed designs of propulsion systems. In doing so. the off-design
`
`
`
`
`
`
`
`
`
`
`operation of a UHBR turbofan is c.':K:In'ti.ned and a design process
`
`
`
`
`
`
`
`
`
`
`for an advanced low-noise propulsion system is demonstrated.
`
`
`
`
`
`
`
`
`The designs an: developed to the point where they can be assessed
`
`
`
`
`
`
`
`
`
`
`
`
`in terms of their performance. weight. and noise. and several pos-
`
`
`
`
`
`
`
`
`
`
`sible engine arrangements arr: presented. Overall.
`this paper
`
`
`
`
`
`
`
`
`makes a contribution to the field of future engine dcsigns for low
`
`
`
`
`
`
`
`
`
`
`
`
`noise and demonstrates the potential of UHBR engines with vari-
`
`
`
`
`
`
`
`
`
`able exhaust systems.
`
`
`
`
`Conu-ibutcd by Iht: lnlcmnttonnl (ins Turbinc Institute of ASME for publication In
`
`
`
`
`
`
`
`
`
`
`
`the lor.!rrNnr_ or Tl.|Ilt58Mn\(?ll!Nl-.ll\'. Mamtscripl rI:ct:ivt:t.t July I3. 2006: final manu-
`
`
`
`
`
`
`
`
`
`
`script rcccivctl July 33. 201'-'|6. Rcvicvv |.‘ttI'KllJL‘l¢IIl by David Wislcr. Paper prtsscntcd at
`
`
`
`
`
`
`
`
`
`
`
`
`
`the ASME Turho Expo EH06: l_nnd, Sort and Air lGT2l|fl(rl. Bart-clona, Spain. Mny
`
`
`
`
`
`
`
`
`
`
`
`
`
`I-ll. Zllflfi. Prtpcr No. GT2ll0(r-9l‘l.‘iS'l
`
`
`
`
`
`
`
`
`Propulsion System Requirements
`
`
`
`The Silent Aircraft is cttpcctctl to use an “all-lifti ng body" style
`
`
`
`
`
`
`
`
`
`
`
`of airframe {[0]. The baseline design has rt payload of 250 pas-
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`Journal of Turbornachlnary
`
`
`
`
`
`Copyright G 2007 by ASME
`
`
`
`
`
`
`
`JULY 2007. Vol. 129 I 479
`
`
`
`
`
`
`
`GE-1009.002
`
`UTC-2011.002
`
`GE-1009.002
`
`UTC-2011.002
`
`

`
`l0EPNdBom.
`
`lSilent Aircraft
`‘target
`
`
`
`Take-offnoise
`
`
`
`Approachnoise
`
`levels for a current 250 paoaartger alreratt oom-
`tho Silent Aircraft target
`
`settgers and a design range of 4000 nautical miles. This mission
`was chosen to give the lowest weight aircraft that would be eco-
`notnically competitive with other civil airlines. A 3-D view of a
`CAD representation of a possible airframe and propulsion system
`is shown in Fig. 2.
`Studies with an airframe design tool were used to detemiine the
`thrust requirements at different points in the mission with corre-
`sponding altitudes and Mitch numbers. The flight mission profile
`was chosen to give the lowest aircraft take-off weight (MTOWI.
`with the assumption that this would minimize the misc radiated at
`take-olT and approach. The methodology and analysis used to op-
`timize the airframe design is described in detail within [l2]. Table
`l summarizes the resulting requirements of the propulsion system
`at key operating conditions in the flight envelope.
`For the purposes of this paper. the noise target of the Silent
`Aircraft is expressed as a peak dBA value that cannot be exceeded
`at any point on the ground outside the airport boundary during
`take-olf and landing in normal operating conditions. This peak
`dBA limit was imposed because this can be linked to both World
`Health Organization guidelines on community noise and data on
`average traffic noise levels in urban areas [l3.l-I]. Normal oper-
`ating conditions are taken as an atmospheric temperature of below
`
`Fig. 2 A 3-D rendering of a candidate Silent Aircraft airframe
`and propulsion ayatorn. taken from [111
`
`Title 1 Propulolort system mission requirements
`Altitude
`Mach
`Total thrust
`Sl-‘C
`number
`1 tn)
`(kNl
`lg/sNl
`
`Noise target
`(peat; dim)
`
`Condition
`
`Sideline
`H yover
`Top-of-climb
`Mid cruise
`Appmaclt
`
`180
`I95
`I2.l92
`I2.570
`I 20
`
`0.23
`0.24
`0.80
`0.80
`0.23
`
`316.8
`172.8
`82.4
`65.4
`<_72.0
`
`—
`--
`—
`
`‘
`
`.
`
`<5”)
`<57.0
`—
`—
`157.0
`
`480 I Vol. 129, JULY 2007
`
`lSA+ I2 K and it runway length of 30(1) m. This allows the air-
`craft to operate in a “nonsilent" mode for any remaining extreme
`conditions (“short & steep." "very hot." and "hot 8: high" take-
`nffl.
`A
`For take-otT. an optimircd departure profile was used in which
`the thrust was managed to achieve the maximum climb rate with-
`out exceeding the noise target in temts of jet noise. This procedure
`is demonstrated in
`and it was found that a total exhaust area of
`l3.2 mi would be required to enable an acceptable departure pro-
`file. The sideline and flyover conditions represent two points in
`the departure profile that are critical in tenns of noise. At sideline.
`the aircraft is still inside the airport boundary and the climb rate is
`highest. At flyover. the aircraft is closest to the population on the
`ground. The sideline lateral position is the same as the lCAO
`certification distance of 450 m. but the flyover point used is closer
`to the runway (4048 in rather than 6500 tn after brakes oft).
`Top-of-climb (DOC) is the condition that detemtines the size of
`the engine. This is where high thrust is required to keep climbing
`and the atmosphere is thin. For an economically viable aircraft of
`the future. the installed engine specific fuel consumption (SFC) at
`cruise was specified to be at least as competitive with the next
`generation high bypass ratio poddcd turhofan engines (lS g/sN
`=0.S3 ll)/lbh). Note that improvements in SFC are beneficial
`in
`terms of total noise because they reduce the weight of fuel that
`needs to be carried. and thus the MTOW.
`For the approach condition. a maximum net thrust target was
`specified in order to limit the airframe drag required. A greater
`drag leads to higher airframe noise through the dissipation of tur-
`bulent kinetic energy in the wake. The minimum thrust specified
`was chosen to be as low as possible while enabling an engine
`spool-u|p time (the time required for the engine to accelerate to
`maximum thrust) that would be comparable to current turbofans.
`Note that all the engine design studies in this paper are matched
`to the same all-lilting body airframe and flight mission. The meth-
`odology applied to develop the engine cycle and the mechanical
`designs should be equally applicable to the propulsion systems for
`other airframe configurations. However. a different airframe or
`installation would have a large impact on the values of many of
`the engine characteristics.
`
`Engine Installation Considerations
`Before the parameters of the engine cycle cant be specified some
`characteftstics of how the propulsion system is packaged with the
`airframe need to be considered. For the Silent Aircraft design. the
`engines are positioned on the upper surface of the airframe. to-
`wards the trailing edge (as illustrated in Fig. 2). This location was
`adopted to take advantage of the perfonnance benefits of bound-
`ary layer ingestion and to maximize the shielding of forward are
`engine noise [4]. It also offers airframe control and safety advan-
`tages. hecause the engines are positioned well behind the passen-
`ger bays [I2].
`A target S-shaped inlet performance was assumed based on
`results in the open literature: for example [l5.l6]. and preliminary
`CFD studies [I7]. Several calculations were completed at
`the
`cruise condition for different numbers of engines and various in-
`take configurations. The current baseline design has 4 separate
`engine units (Fig. 2), which gives an acceptable fan diameter and
`good installation performance. The mesh geometry and Mach
`number contours. from a calculation of this configuration. are as
`shown in Figs. 3 and 4. respectively. Figure 4 shows how a large
`region ofhigh loss flow builds up at the bottom ofthe inlet duct as
`the engine face is approached. and this is typical for an S-shaped
`duct
`The final propulsion system for the Silent Aircraft is expected
`to use boundary layer ingestion (EU) to give a fuel bum benefit.
`as mentioned above. and as discussed in [2]. BL] introduces sig-
`nificant challenges to both the engine and airframe design.
`In
`panicular. BLI generates additional nonunifonn flow distortion in
`both the radial and circumferential directions. which is present at
`
`Transactions of the ASME
`
`GE-1 009.003
`
`UTC-2011.003
`
`

`
`
`
`
`
`Fig. 3 View at the surtaoo mesh for tr louronglno Installation,
`talten from [17]
`
`all flight conditions. BLI also changes the engine thrust require-
`ments because any boundary layer fluid that passes thmugh the
`engine and contributes to thrust would have otherwise contributed
`to airframe drag. Thus. to progress the engine design for the stud-
`ies in this paper. the engine inlets were temporarily assumed to be
`boundary layer divening (BLDl. This enables the engine thrust
`requirements to stay as per Table I. The impact of BLI on the
`engine performance and the airframe requirements will be in-
`cluded in future work.
`CFD results for S-shaped inlet ducts all show significant sepa-
`rated regions at the fan-face. From the inlet highlight to the fan-
`face. at typical S-shaped inlet was found to have a pressure recov-
`ery (pm/pm) of about 0.96. This value can be applied to both BL]
`and BLD cases, and it was used in the following engine design
`studies as a target performance for the engine inlet. Overall. it is
`expected to be a lower-bound estimate because design improve-
`ments and flow control should be able to reduce the losses. The
`level of circumferential distortion was also determined from the
`predictions. and in terms of DC60 (an industry measure of the
`severity of How nonuniformity). the S-duct gave values of around
`20%. The impact of this distortion on the system design will be
`explored in detail in future research. because it is mainly a t:on—
`sequence of BLl. Tlx: designs presented in this paper are therefore
`intended to tolerate this level of distortion. but are not optimized
`for performance with it present.
`The engine exhaust is considered as a long cylindrical duct in
`which the core and bypass streams are mixed completely. fol-
`lowed by a loss-free variable nozzle. The exhaust duct pressure
`losses were determined using simple compressible pipe flow
`analysis lFanno line flow) with skin friction coefficients appropn'-
`
`
`
`Flg. 4 Contours of their number through a four-engine Instal-
`lation. taken from [17]
`
`Journal or Turbomachlnory
`
`
`
`Fig. 5 scnernetlc of eng
`adapted from [18]
`
`Inc layout snowing station numbering.
`
`ate for the surface of a perforated acoustic liner. The exhaust duct
`size was set to match the maximum nozzle area required for a
`quiet take-off. This size of duct was used to avoid a "diffusing
`nozzle" being necessary at airy point
`in the aircraft flight enve-
`lope. The length of mixer duct was set at 2 fan diameters to
`accommodate a large area of downstream acoustic liners. This led
`to an exhaust pressure recovery (pm/pm.) of 0.98 for the designs
`in this paper.
`
`Engine Cycle Design
`The engine configurations developed in this paper are ultraltigh
`bypass ratio (UHBR) turbofans combined with variable area ex-
`haust nozzles. This configuration was identified in [Z]. and for the
`aircraft mission requirements it was expected to be more suitable
`than other possible variable cycle systems such as bypass stream
`ejcctors or a system with auxiliary fans. The optimum solution for
`a different mission requirement may be quite different. The engine
`station numbering used for the thermodynamic cycle is as shown
`in Fig. 5.
`To allow for technological advances. 2025 estimates of peak
`component efficiencies and metallurgical limits were made by ex-
`trapolating historical trends. These were imposed as limits on the
`engine cycle temperatures and component eflicicncics that could
`not be exceeded at any point in the engine operation. It was ex-
`pected that it future quiet engine would have a similar maximum
`fan capacity to today's turbofan designs. but that this could be
`achieved at a lower fan pressure ratio and tip speed (see [2]). A
`maximum fan capacity was therefore imposed as a constraint and
`combined with a generic low pressure ratio fan characteristic to
`estimate off-design performance variations. It was also predicted
`that advances in mechanical propenies would allow the hub-to-tip
`radius ratio of a future fan to be lower than current designs. A
`value of 0.25 was used to minimize the fan diameter (current
`designs are typically in the range 0.3-0.35l.
`In order to develop an engine cycle. a design condition is cho-
`sen to fix the engine size and key parameters. For this study. the
`top-of-climb point was used and the themrodynamic cycle was
`optimized to minimize the fan diameter and fuel consumption at
`this condition. The top-of-climb point was then considered with
`the other off-design conditions and some iteration was employed
`to optimize the perfonnance for every engine operating condition
`in Table l.
`
`Cycle Optlmlntion at Top-of-Cllmb. The engine design cycle
`was developed using Ga.sTurbl() [18] with the aim of producing
`the most compact and fuel-efficient engine that would satisfy the
`requirements in Table I. Figure 6 shows how an engine cycle
`appropriate for the future Silent Aircraft was evolved from a cur-
`rent conventional
`turbofan. Each bar in the figure represents a
`redesign of the engine in which the fuel consumption has been
`minimized and the net thrust and temperature limits have been
`constrained. The relative heights of the adjacent bars show the
`impact of each design change on engine fuel consumption and
`engine size. The aim of such a chart is to show that
`the final
`
`e€i'1’ti3’§.ii’bK"
`
`’ “"
`
`UTC-2011.004
`
`

`
`:— T
`-_
`I LPC lnletTip Diameter
`D Specific Fuel Consumption
`
`3
`
`'
`
`I
`
`I
`‘
`
`.-s
`E
`I.
`3
`g 2.4
`.9
`‘U
`g 2.2
`‘L
`
`3.4
`
`3.2
`
`3
`
`2.3
`
`2.6
`
`2
`
`L8
`
`i.6
`I 4
`
`E
`3
`S
`l-
`2
`3
`g
`U
`ix
`2
`E
`5
`B
`‘
`I
`¢
`..
`Q
`8
`'-
`-
`‘
`'3
`U
`>
`U
`
`E
`I
`
`Z
`3
`5
`3
`9
`-
`-l
`2
`V
`“‘
`Q
`3
`:
`<
`m
`i-
`5'5
`E
`z
`g
`\fl
`U
`Q
`3
`8
`2
`
`vi
`~
`2
`tu
`
`3
`“
`9
`-
`:
`Q4
`..
`9
`8
`"
`2
`in
`0
`"'
`:
`n
`T.’
`E
`3
`
`-
`3
`5:
`..
`0
`E
`_,
`"
`*3
`3
`o
`0-
`0
`3
`‘
`w
`g
`
`7
`
`g
`5
`a
`-H
`3
`u
`I
`3
`O
`"
`O
`In
`2
`3
`2
`6
`
`U‘
`.n
`ll
`-'
`at
`a
`:
`:2
`D
`o
`°~
`'2
`"
`It
`:
`:
`
`4
`
`in
`9
`ll
`-—'
`n
`m
`:
`:2
`:
`o
`3'
`.
`T
`I
`:
`:
`i
`S
`
`II!
`5
`in
`‘2’
`-
`at
`t
`in
`at
`E
`.
`I-
`S
`E
`8
`
`>
`V
`E,
`Q
`It
`‘
`In
`g
`2
`9
`
`
`
`0.62
`
`0.6
`
`0.53 E
`<
`D
`0.56 :
`C
`-8
`0.54 Q
`E
`0.52 V‘
`C
`0
`1
`V
`(1)
`.2
`u
`0.43 E
`u
`8.
`0.46 tn
`
`0.5
`
`0.44
`0 42
`
`8
`5
`-I
`:
`u
`>
`3
`E
`3
`_,
`_
`3
`3
`
`3
`3
`'3
`I
`ill
`0
`:
`..
`4
`5
`O
`E
`‘A
`14
`
`N
`5
`—
`*
`3
`3
`3
`
`3
`u
`"
`U
`S
`3
`3
`l2
`
`IO
`
`ll
`
`‘I3
`
`Fig. 6 Evolution of the Silent Aircraft onglno cycle at top-of-climb assuming a tour-engine system
`
`Design step
`
`design cycle is feasible. The changes in the height of the bars
`between each design show the incremental effects of changing the
`design cycle paraineters.
`the design
`The first
`three steps‘ shown in Fig. 6 represent
`changes necessary to match toi.la_v's turhofan engine to the top-oll
`climb condition in Table l. The two subsequent changes indicate
`how a large performance benetit can be attributed to the low fan
`pressure ratio that is specified (steps -1 and 5). A drawback of this
`is that the fan diameter increases significantly as l-‘PR is reduced.
`Higher temperature limits and improved turbine performance con-
`tribute significantly to improving core efficiency (steps 6. 7. and
`8). However. as the engine efficiency is improved. the fun diam-
`eter has to increase to maintain the same net thrust. This effect is
`also seen when the fan efficiency is increased and the l()\\c.\ due
`to intemztl air systems are reduced (steps 9 and l(ll, The exhaust
`duct that is specified is larger than optimum to match the variable
`area nozzle (step II) and this also leads to a slightly larger fan.
`The cycle was then optimized for best
`thermodynamic perfor-
`mance tstep I2). This involved iterating to detemiine the bypass-
`to-core total pressure ratio (p(,,.,/pm) that gave the minimum SEC.
`The use ofthe high capacity. lo» huh-tip-ratio fan (step lfll gives
`further improvements. Howc\er.
`the introduction of the S-duct
`inlet total pressure loss (step I4) increases the fan diameter and
`significantly worsens the overall performance.
`Note that the design cycle used in this paper has a fan pressure
`ratio of L45, and for a four-engine case. a fan diameter of 2.l6 m.
`The choice of design FPR is a compromise that is driven by Se\-
`eral factors. A lower value leads to higher propulsive efficiency at
`the cost of [1 larger engine size. which increases the total installed
`drag. in terms of noise. as FPR reduces it becomes easier to meet
`the jet noise target. and the nozzle area change needed between
`take-oil‘ and top-of-climb is minimized [I9]. Fan source noise also
`tends to reduce vtith FPR, as shown later in this paper in Fig. I4.
`Unfortunately. a lower FPR design is heavier and more sensitive
`to inlet distonion and to iiistalliition pressure losses. as shown in
`[2]-
`
`482 I Vol. 129. JULY 2007
`
`A top-of-climb l-‘PR of" I.-l5 was therefore chosen as the lowest
`possible value that would be achievable with it robust mechanical
`design. The corrtsporiding engine bypass ratio is
`I5 5. which
`clearly makes it a IFHBR. However. BPR is not a good design
`parameter to characterize the engine because it changes signifi-
`cantly between operating conditions (see Table 2. later). The SFC
`at top-of-climb is 14.7 g/5N. which is slightly better than the best
`turbofars operating today. This seems realistic for a UHBR engine
`in 2025 within an S—type inlet.
`
`Off-Design Operation. Using the final cycle design developed
`above. the engine parameters at other points in the flight mission
`were determined. With the engine size fixed and the thrust con-
`strained the main degree of freedom available is the nozzle set-
`ting. At each of the llight conditions in Table I. the fan can oper-
`ute aiiywhen: along a cltaractcristic of cunstatit
`lhntst. Figure 7
`shows scaled constant-speed fan characteristics based on [20] with
`constiiiit-tliruut characteristics overlaid (dashed) for each of the
`key operating conditions. The optimum operating points used for
`the tinal design are marked as small circles. The precise perfor-
`
`Tablo 2 Cycle parameters for the Silent Aircraft engine design
`In this paper
`
`Parameter
`l-‘PR
`M1 \ {M
`M,
`M ,
`,,,:h'
`TH
`Tm (rgri
`OPR
`BPR
`SK‘
`
`Sideline
`l .27
`90
`+35
`0.64
`94 5
`‘llll
`I730
`-Al 0
`l9.U
`8.8
`
`Top-oflclimb
`L45
`I00
`0
`on!»
`90.4
`900
`ltillltl
`57.4
`|S.5
`l-1.7
`
`Cniise
`L40
`99
`+3
`n70
`93.4
`340
`1700
`53.9
`lb.8
`l4.2
`
`L'nit
`—
`G;
`‘7'
`_
`1;;
`K
`K
`—
`-
`g,l.\N
`
`TIIHSGCIIOTIS Of "'08 ASME
`
`GE-1 009.005
`
`UTC—20l1.005
`
`
`
`

`
`
`
`
`
`.
`0.0
`0.0
`0.7
`0.0
`rrrtffolpo reldivo to design punt
`
`1
`
`1.1
`
`.-‘0
`
`.“5,
`
`FanProaaureRatio -.1
`
`0.4
`
`0.5
`
`Fig. 7 operation at the Silent Alrcratt onglno tan tor a variable
`nozzle design
`
`Fig. 9 variation In tan Mtlcloncy. lot noise, and tan tip rolatlvo
`Iuetr number with nozzle area.
`
`mance depends on the shapes of the fan characteristics. so these
`plots can be viewed as an example case. Improved fan character-
`istics for a full 3-D fan design are developed in the companion
`paper [I9].
`Tire characteristics in Fig. 7 have been scaled around the design
`FPR of L45 (lop—of—climbl and a peak rotor isentropic efficiency
`of 945'} (the maximum possible expected in W25). By varying
`the nozzle exit arm while maintaining the net thmst constant-
`thrust lines could be produced. These were further constrained by
`shaft speed and temperature limits. The optimum top-of-climb
`point is positioned towards the stability margin on the l(X)% speed
`characteristic. This was done primarily so that the exhaust nozzle
`could be opened sufticienuy at sideline to give higher tan capacity
`at this condition (and thus low jet noise). while keeping high
`efiiciency. The design condition can be positioned further down
`the maximum speed characteristic. but this reduces the operating
`range available at other conditions. The fan capacity at cruise is
`allowed to increase slightly without exceeding the design fan shaft
`speed to give improved efficiency.
`Figure 8 shows the fan characteristics if the same mission re-
`quirements and design engine cycle are assumed for a hired nozzle
`
`4
`
`engine. In this case. the fan is constrained to work at I single
`working point for each flight condition. The top-of-climb point is
`positioned so that the cruise condition is at peak efficiency. The
`sideline. flyover. and approach points are thus fixed at lower flow
`rates. which are at higher fan pressure ratios and closer to insta-
`bility than for the case with a variable nozzle.
`Table 2 details the key cycle parameters at sideline. topol-
`climh. and cruise corresponding to the operating points in Fig. 7.
`This illustrates the component performances and the cycle tern-
`peratures that are required to achieve the design requirements. It is
`important to emphasize that the operation of the Silent Aircraft
`engine with a variable area nozzle differs significantly from that
`of an engine with a fixed nozzle operated for a conventional air-
`craft. Firstly. the fan pressure ratio at take-oft is much lower than
`at cnrise or top-of-climb. The principal reason is that only a frac-
`tion ot the available thrust at take-off is needed. The total sea level
`static thrust available from the propulsion system (all engines) is
`about 570 kN. and less than 60% of this is needed at the sideline
`condition. The low thnrst requirement at sideline is key to mini-
`mizing the jet noise. and this is further exploited with an opti-
`mized talteoff procedun: that is described in [3].
`Another unusual aspect of the design is that the fan speed is
`similar at all the three conditions in ‘lhble 2. and the fan-face
`Mach number is consistently high. Previous studies [2] showed
`that a high fan capacity at take-off leads to lower jet noise and
`Figs. 7 and 9 demonstrate how this can be achieved with a vari-
`able exhaust nozzle.
`The variations in cycle temperatures and pressures are also dif-
`ferent from a conventio-nal turbofan. Usually the cycle tempera-
`tures are all highest at take-ofi". and it is this condition that is most
`demanding in terms of the mechanical stresses. For the design
`developed here the compressor outlet
`temperature is highest at
`take-ofi‘. but only slightly above the top-of climb and cruise
`points. The turbine entry temperature is a maximum at rop-ot'-
`climb. where the overall pressure ratio is also much greater than
`the sideline condition. This occurs because the thrust requirement
`of the engine during take-off is only :1 fraction of the total thrust
`available.
`_
`The variable nozzle has the potential to reduce fuel consump-
`tion. It should be possible to carefully control the nozzle position
`to maximize the fan efficiency at all flight conditions. as indicated
`in Fig. 7. For a fixed nozzle design. a fan is constrained to operate
`on 11 worlting line that might not be at peak efficiency. In addition.
`the fan characteristics of an engine operating in-service may not
`be exactly as predicted. A variable nozzle enables performance
`discrepancies to be conected during flight. ensuring the optimum
`
`JULY 2007. Vol. 129 I 403
`
`G E-1 009.006
`
`UTC-2011.006
`
` laentropreefliereneyoO_a8'08 .\§1*-.
`
`/‘\
`
`\
`
`\\
`
`,
`
`\\
`
`/WI
`/ /1
`/A
`1/
`‘ //
`2/
`I///
`
`"
`
`
`
`0 3.4
`
`05
`
`oh
`ofa
`0.7
`ofe
`rm"l'°Ip°relntivab design pair!
`
`1
`
`1:1
`
`1.:
`
`.
`
`9 1.5-
`1:
`
`1,4
`
`~ .
`it
`;.~-
`-
`_,a'_'_’‘: TT T3‘: I
`,/
`\\ A
`.
`,.v"
`.
`I
`Na
`“
`'2 srrI‘g€.“.'3.-g...‘ ___\ %"
`* K - Mover
`"
`C
`‘A
`\
`N.Nl’a.1lncreasing
`.
`_ Qua
`‘L T.
`|
`‘
`t
`A
`l
`A
`A
`.4
`05
`on
`0.7
`0.3
`on
`1
`1.1
`
`1.2
`
`mx"I'DIpo rehtiveto design poirt
`
`Fig. 0 Operation of the Silent Aircraft englno tan tor a fixed
`noulo doalgn
`
`Journal ot Turbomachlnory
`
`

`
`‘title 3 Principal mechanical parameters for the engine dc-
`clgns presented
`
`Design A
`
`Design B
`
`Design C
`
`Design D
`
`Three-spool Two-spool Twrrspool. Multiple fan
`turbofan
`geared fan
`slower fan
`system
`2.l6
`Z.l6
`2.l8
`L2!
`3.4!:
`2.42
`2.70
`2.70
`
`4
`20
`
`4
`20
`
`—
`
`— —»»ri=x':4.-
`,
`
`Conliguruion
`
`1), (ml
`1..., (ml
`"W
`Number of fun
`rotor blades
`
`Fig. 10 Doolgn A: 3-spool conventional turbofan design for
`the Sllont Alreralt
`
`the configmations satisfy the cycle
`presented in this paper. All
`parameters shown in Table 2 and the mission requirements in
`Table I.
`Design A is a conventional three-shaft turhofan architecture.
`The general arrangement for this design (Fig. 10) was obtained
`using current design levels of aerodynamic loading and stress. and
`typical geometrical constraints for the turbomachinery annulus.
`There are several probletm with this design that make it an unre-
`alistic solution. Firstly. the LP turbine has nine stages. making it
`very bulky. heavy. and noisy. This is necessary in order to drive
`the relatively large fan at
`low rotational speed. The low shalt
`speed also leads to high torque. dcrrtarrding very thick shafts. The
`core annulus is quite convoluted and S-ducts with dramatic
`changes in radius between the IP and HP turbines are required.
`Ten stages of HP compressor are required to achieve the cycle
`OPR and this demands a blade height in the final stage of less than
`I0 mm. This size of blade would sulfer significant losses from
`Reynolds number effects and tip clearance flows. and it would he
`very difficult to manufacture accurately with current tools.
`Design B. illustrated in Fig. ll, was developed in order to ad-
`dress the problems identified in Design A. To reduce the LP tur-
`bine size a 3:! reduction gearbox has been placed between the fan
`
`Fig. 11 Design 8: Geared turbotan tor the sllent Alreratt with
`artlal-radial HP compressor
`
`eEJlll‘6'§'.'ll‘6?’ "" ‘s"‘
`
`UTC-2011.007
`
`380
`
`330
`‘I
`10
`I0
`
`Max. M... (nus)
`IPC/booster stages
`HPC stages
`HPC min. blade
`height (rnntl
`9
`LP turbine stages
`
`w,,, ta») 100
`
`efficiency is achieved.
`A variable exhaust nozzle can also improve the engine operabil-
`ity. During take—off. with the male fully opened. the fan operates
`well aw