Breakdown of aerodynamic interactions for the lateral rotors on a compound helicopter

Breakdown of aerodynamic interactions for the lateral rotors on a compound helicopter

Stokkermans, Tom; Veldhuis, Leo; Soemarwoto, Bambang; Fukari, Raphaël; Eglin, Paul

DOI

10.1016/j.ast.2020.105845

Publication date

2020

Document Version

Final published version

Published in

Aerospace Science and Technology

Citation (APA)

Stokkermans, T., Veldhuis, L., Soemarwoto, B., Fukari, R., & Eglin, P. (2020). Breakdown of aerodynamic

interactions for the lateral rotors on a compound helicopter. Aerospace Science and Technology, 101,

[105845]. https://doi.org/10.1016/j.ast.2020.105845

Important note

To cite this publication, please use the final published version (if applicable).

Please check the document version above.

Copyright

Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons. Takedown policy

Please contact us and provide details if you believe this document breaches copyrights. We will remove access to the work immediately and investigate your claim.

This work is downloaded from Delft University of Technology.

Contents lists available atScienceDirect

Aerospace

Science

and

Technology

www.elsevier.com/locate/aescte

Breakdown

of

aerodynamic

interactions

for

the

lateral

rotors

on

a

compound

helicopter

✩,✩✩

Tom Stokkermans

a

,

∗

,

Leo Veldhuis

a

,

Bambang Soemarwoto

b

, Raphaël Fukari

c

,

Paul Eglin

c a_{DelftUniversityofTechnology,Delft,theNetherlands}

b_{NetherlandsAerospaceCentreNLR,Amsterdam,theNetherlands} c_{AirbusHelicopters,Marignane,France}

a

r

t

i

c

l

e

i

n

f

o

a

b

s

t

r

a

c

t

Articlehistory:

Received6August2019

Receivedinrevisedform17February2020 Accepted31March2020

Availableonline28April2020 CommunicatedbyGeorgeBarakos

Auxiliaryliftand/orthrustonacompoundhelicoptercanintroducecomplexaerodynamic interactions between the auxiliary lift and thrust components and the main rotor. In this study high-fidelity computationalfluiddynamicsanalyseswereperformedtocapturethevariousaerodynamicinteractions whichareoccurringfortheAirbusRACERcompoundhelicopter,featuringabox-wingdesignforauxiliary lift incruiseand wingtip-mountedlateral rotorsin pusherconfiguration forauxiliary thrust incruise and counter-torque in hover. Althoughthe study was limited to aspecific geometry, the effectsand phenomena are expected to be to someextent applicable in general for compound helicopters and wingtip-mountedrotorsinpusherconfiguration.Aquantitativeindicationoftheaerodynamicinteraction effects could be established by leavingaway different airframe components in the simulations. The downwash ofthe main rotorwas found to causea smallnegative angle of attack in cruisefor the wingsandlateral rotorsand impingeddirectlyonthe lateralrotorsinhover,resultinginmoderate to verysignificantsinusoidallyvaryingbladeloading.Thewingincreasedlateralrotorpropulsiveefficiency incruisethroughitswingtiprotationalflowfieldand toalesserextentthroughitswake.Anupstream effectofthelateralrotorsonthewingloadingwasalsofound.Inhoverthewingcausedanetincrease inleftlateralrotorthrustasthedeflectionofthemainrotorflowtowardstherotorresultedinalocal thrustdecreaseandthelowmomentuminflowtotherotorfromthewakeofthewingresultedinalocal thrustincrease.Asmallthrustdecreasefortherightlateralrotorwasfoundduetothewingdisturbing itsslipstreamasthisrotorproducedreversedthrust.Ingeneral,verysignificantaerodynamicinteraction effectscanbeexpectedwhenamainrotor,lateralrotorsandwingareinproximitytoeachother.

©2020TheAuthor(s).PublishedbyElsevierMassonSAS.ThisisanopenaccessarticleundertheCCBY license(http://creativecommons.org/licenses/by/4.0/).

1. Introduction

Thereis agrowing needforaircraft withverticaltake-off and landingcapabilitythatcangofastandfar[1].Whilehelicoptersare excellentforverticaltake-offandlanding,theyonlyhavealimited maximumspeed:Theasymmetricflowconditionofthemainrotor athighspeedcausescompressibilityeffectsontheadvancingblade sideandstall onthe retreatingblade side ofthemain rotor that limit its liftingandpropulsive capability [2,3].A compound

heli-✩ _{Funding:ThisprojecthasreceivedfundingfromtheCleanSky2Joint} Undertak-ingundertheEuropeanUnion’sHorizon2020researchandinnovationprogramme undergrantagreementNo.685569—PROPTER—H2020-CS2-CFP01-2014-01. ✩✩ _{Thispaperreflectsonlytheauthor’sviewandtheJUisnotresponsibleforany} usethatmaybemadeoftheinformationitcontains.

*

Correspondingauthor.

E-mailaddress:T.C.A.Stokkermans@tudelft.nl(T. Stokkermans).

copterovercomesthemainrotorlimitation athighspeedby aux-iliaryliftandthrustdevices[1,2].Anexampleofsuchacompound helicopterconcept is theAirbus RACER(Rapid And Cost-Effective Rotorcraft) [4,5],optimized forahighcruise speed of220kts. In Fig.1asketchisshownofthehelicopter,withoutmainrotorand tailplanes. Supporting the main rotor, auxiliary lift in the cruise condition isprovided by abox-wing, whilewingtip-mounted lat-eralrotorsinpusherconfigurationgenerateamajorportionofthe requiredthrust.Furthermore,theselateralrotorsprovide counter-torque in the hover condition to balance the main rotor torque. In the cruisecondition, thevertical fins (not shown)produce all thecounter-torque.Theyarealsofittedwithflapsinorderto ad-justthecounter-torqueasfunctionofe.g.airspeed,airdensityand mainrotortorque.Forbrevity,fromthispointon,lateralrotorswill bereferredtoaspropellers.Comparedtoasinglewingdesign,the box-wingdesignreducestheoverallsurfaceaffectedbythe down-washofthemainrotorinhover,whiledeliveringtherequiredlift

https://doi.org/10.1016/j.ast.2020.105845

Fig. 1. SketchesoftheAirbusRACERwithoutmainrotorandtailplanes.PropellerbladephaseangleϕandaxialandtangentialvelocitycomponentsVaandVtareindicated.

in cruise. It can also improve lift-over-drag ratio compared to a planarwingdesign[6].

Becauseofthecloseproximityofthebox-wingandmainrotor, thepropellersexperiencevarious aerodynamicinteractions. These interactions significantly differforeach flight condition. The pro-pellerinstallationeffects forthiscompoundhelicopterhavebeen described by Stokkermanset al. [7] andto some extentby Wen-trupet al. [8]. Furthermore,theinteraction ofa wingtip-mounted pusher-propellerwiththeupstreamwinghasbeenstudiedfor pla-nar wings by Refs. [9–12]. The extensive experimental study of BainandLandgrebe [13] coversmainrotor–wingaerodynamic in-teraction andmain rotor–propeller aerodynamic interaction fora compoundhelicopter, treatingboth casesseparately.The focusof that study was ontheblade moments whichthemain rotor and propellerexperience. Mainrotor–winginteractionwas also inves-tigatedby Lynn [14] andspecificallyfortheRACER,thedesign of thebox-wingswithmainrotor interactioniscoveredbyWentrup et al. [8].

Inthereviewoflessonslearnedfromthecompoundhelicopter studies done by NASA andthe US Army by Yeo [1], the need is stressedforhigh-fidelitycomputationalfluiddynamicsanalysesto capturethevariousaerodynamic interactionswhichare occurring forcompoundhelicopters.AccordingtoYeo [1],theseaerodynamic characteristics could then be used for calibration of lower order models,andlayout andshape refinement.Inlinewiththisneed, aspart of the Clean Sky 2 PROPTER project(Support to aerody-namicanalysisanddesignofpropellersofacompoundhelicopter), inthis articlea high fidelity breakdownis presented ofthe var-ious aerodynamic interaction effects between the propellers, the box-wingsandthemainrotordownwash.Theseeffectsinclude lo-calchangesindynamicpressureandangleofattackfromvortices, boundary layers and rotor momentum sources. The focus of this articleisontheresultingchangesinloadingandefficiencyofthe propellersandliftanddragofthebox-wing.

2. Computationalsetup

Theaerodynamic interaction effectswere investigatedthrough a series of Reynolds-averaged Navier-Stokes (RANS) CFD simula-tions. InFig.2 thevarious simulatedconfigurations are depicted. Byleavingawaydifferentairframecomponentsinthesimulations, a quantitative indication of the aerodynamic interaction effects couldbeestablished.Themodelwas simplifiedbyremovalofthe tailunitandthetime-averagedeffectofthemainrotorwas intro-ducedbyanactuatordiskimplementation,similartoe.g.Batrakov et al. [15],whichconsistsofradiallyandcircumferentiallyvarying momentumandenergyjumpconditionsbasedonprovidedblade loadingdistributions.Notethatrotorandrotor-headwake interac-tionswiththetailunitandresultingdesignoptimisationofthetail

Fig. 2. Thefivedifferentsimulatedconfigurationsincludinglabels(P:propeller;N: nacelle;W:wing;F:fuselage;R:rotor).

unithavebeenstudiedseparatelybyLienardet al. [5] andSalahel Dinet al. [16] respectively.

Adaptive mesh refinement techniques such as described by Öhrle et al. [17] wouldenable thefullsimulationofthemain ro-tor withreduced computational cost, butforthis projecta large quantity of simulations were neededto study the interactions in the various configurations and operating conditions, necessitat-ing themorecomputationalcost effectiveactuatordiskapproach. Of course, one should consider that this setup neglects any of the transient effects that the main rotor blade tip vortices and wakeshaveonthepropellerloadingase.g.describedbyThiemeier et al. [18]. Furthermore,no interaction effect ofthe propeller on the main rotor loading is present asthe actuator-disk loading is prescribed.Asisknownfromtheextensiveexperimental investiga-tionofBainandLandgrebe [13],apropellerclosetothemainrotor mayaffecttherotorflappingamplitudeandbendingmomentsdue toitspressurefield,andassuchthemainrotorloading.However, theactuator-diskloadingincludestheasymmetryoccurringinthe loading onthe mainrotor in forwardflight betweenthe advanc-ingandretreatingbladesideandtheradialnon-uniformitypresent intheloadingonrotorsandassuchrepresentsthetime-averaged effectofthemainrotorontheflowfieldaccurately.

2.1. Flowsolver,domainandboundaryconditions

The flow solver used for these simulations was ENFLOW, a multiblock structured solverby NLR[19–22].The RANSequations were discretized in space by a second-order cell-centred finite-volume method,using central differencesand artificial diffusion.

Fig. 3. Comparisonofthecomputationaldomainforthecruiseconditionand en-largeddomainforthehovercondition.

The steady flow equations were solved by a multi-grid scheme, using as relaxation operator a Runge-Kutta scheme. For tran-sient solutions, the flow equations were integrated in time by a second-order implicit scheme, using the dual-time stepping method.A timestep equivalent to1 deg ofpropellerrotation was usedforthefinalresultsascommonlyfoundinpropellerresearch [23,24]. To model turbulence in the cruise condition, an explicit algebraicReynolds-stresswasselectedfortheReynolds-stress ten-sor(EARSM), combined withthe TNT k

−

ω

modelto determine therelevant turbulenttime andlength scales [25]. Forthe hover conditionthek

−

ω

SSTmodelwasusedtopromoteconvergence ofthiscomplex flowfield[26].ForthePNconfiguration,transient simulations were performed whenever the propeller was at an-gleofattack.Forzeroangleofattack,a singleblade passagewas modelledinasteadysimulationwithcyclicperiodicboundary con-ditionstoreducecomputationalcost.

ForthePNWFR,PNRandPNWFconfigurations,transient simu-lationswereperformed,whiletheNWFRconfigurationwas simu-latedsteadytoreducecomputationalcost.Thedomaindimensions for thesesimulations are depicted in Fig. 3. For the cruise con-dition, the dimension in the freestream direction L was chosen larger than the other dimensions to diminish the effect of the boundaryconditionsontheflowfieldnearthehelicopter. Further-more,onall boundaries,generalfree streamboundaryconditions wereprescribedbasedonRiemanninvariants.Thefreestream air-speedinthe cruisecondition was setto 220kt (113

.

2m

/

s) with a statictemperature of 276

.

3K and densityof 1

.

024kg

/

m3. For thehovercondition,a largerdomainwasrequiredtoprevent un-wantedrecirculation andthus unwanted influence ofthe bound-ary conditions on the flowfield near the helicopter. The bottom boundaryconditionwasplacedatadistanceequaltothehover al-titudehhoverandmodelledasaslipwall.Averysmallfreestream velocitywas prescribed tofurther reducerecirculation inthe do-mainandthestatictemperaturewassetto306

.

2K anddensityto 1

.

112kg

/

m3_.

Transientsolutionsweresoughtafterwhichwereperiodicwith thepropellerbladepassagefrequency.Theproceduretoobtainthis periodicsolution differedforeach flight condition. Forthecruise conditionthefollowingprocedurewasappliedoverthreegrid lev-els(coarse,mediumandfine):

1. Steadycoarsegridsimulationfor1,000iterations.

2. Continuewithtransientcoarsegridsimulationfor4propeller rotations.

3. Continue with transient medium grid simulation for 3 pro-pellerrotations.

4. Continue with transient fine grid simulation for 2 propeller rotations.

Thehoverconfigurationrequiredadifferentrecipetoavoid di-vergence in the solver and to reach periodic behaviour with the blade passage frequency.It was found that toreach periodic be-haviour for this flight condition, many propeller rotations were requiredastheflowfieldisinducedbythemainrotoractuatordisk andpropellers.Alargertimestepequivalentto10 degofpropeller rotationwasusedtoreachtheperiodicbehaviourwitharelatively lowcomputationalcost.Eventuallyasmalltimestepequivalentto 1 degof propeller rotation was againused to obtainthe desired temporalresolution:

1. Steadycoarsegridsimulationfor10,000iterations.

2. Continuewithsteadymediumgridsimulationfor4,000 itera-tions.

3. Continuewithsteadyfinegridsimulationfor2,000iterations. 4. Continue withtransient fine grid simulationfor30 propeller

rotationswith10 degequivalenttimestep.

5. Continue with transient fine grid simulation for 4 propeller rotationswith1 degequivalenttimestep.

To obtain time-averagedresults, a final blade-passage (60deg ofpropeller rotation) was simulated,over whichall quantities of interest were averaged.The computationalcost ofa PNWFR con-figurationsimulationwas approximately180hr on 240 coresofa modernHPCclusterforthecruiseconditionandthedoubleforthe hovercondition.

2.2. Gridandgriddependencystudy

Blockstructuredgridswereconstructedandthepropellersand spinnerwere in aseparate domain fromtherest oftheairframe to allow propeller motion witha sliding mesh interface. A slid-ingmeshinterfacewasalsopresentaroundthepropellerbladeto allow changeof propeller bladepitch without redefinitionof the structuredgrid.Thefinegridcontainedabout153 millioncellsfor thecruiseconditionand163 millioncellsforthehovercondition. Thegridsizeofthehoverconditionwas slightlydifferentasa re-sultoftheenlargeddomainandtheneedforadifferentfirstlayer thickness.Inordertocomplywiththeturbulencemodels,the di-mensionlesswalldistancey+onallno-slipwallsofthemodelwas lessthanone.

Inordertoassessthedependencyoftheresultsonthegrid,the grid convergence index(GCI) was evaluated. The procedure from Celik et al. [27] wasfollowed.Fora quantityofinterest,

φ

1 isits valueonthefinegrid,

φ

2onthemediumgridand

φ

3onthecoarse grid.Themediumandcoarsegridwereobtainedbystructuredgrid coarsening.Theapparentorderoftheschemep based onthe so-lutionofthequantityofinterestonthethreegridsisgivenby:

p

=

1 ln

(r

21

)

|

ln

(

ε

32

/

ε

21

)

|

with

ε

32

/

ε

21

=

φ

3

− φ

2

φ

2

− φ

1 (1)

where r21 is the grid refinement ratio from the medium to the fine grid, 2 in this case. Note that this procedure only works if theratioof

ε

32

/

ε

21ispositive.Negativevaluesindicateoscillatory convergence. The fine-grid convergence index with a 1.25 safety factorisgivenby:

GCI21_fine

=

1

.

25e

21 a r₂₁p

−

1 with e 21 a

= |

φ

1

− φ

2

φ

1

|

(2) withe21

a theapproximate relative error.Since the grid isrefined

in a structured manner and the solver is second order accurate, alsoanerrorbasedonRichardsonextrapolationcanbecalculated, wherep

=

2 inthepreviousequations.Thefinegridrelativeerror basedonRichardsonextrapolationis[28]:

Table 1

GridconvergencestudyresultsinthecruiseconditionbasedGCIand Richardsonextrapolation.

left propeller right propeller airframe thrust power thrust power lift drag

ε32/ε21 9.2 30.4 7.9 12.4 5.2 4.3 p 3.2 4.9 3.0 3.6 2.4 2.1 GCI21 fine(%) 0.1 0.0 0.2 0.0 1.4 6.9 E1(%) 0.2 0.0 0.4 0.1 1.6 6.2 Table 2

GridconvergencestudyresultsinthehoverconditionbasedGCIand Richardsonextrapolation.

left propeller right propeller airframe thrust power thrust power lift drag

ε32/ε21 3.0 6.6 8.3 7.9 -1.4 1.9 p 1.6 2.7 3.1 3.0 0.5 1.0 GCI21 fine(%) 0.7 0.3 0.1 0.2 -∗ 44.2 E1(%) 0.4 0.5 0.1 0.4 2.3 11.1 ∗_{oscillatoryconvergence.} E1

=

e21_a 22

₋

₁ (3)

Theadvantageofthislateruncertaintyestimateisthatitdoesnot dependonthecoarsegrid,whichmightbeout oftheasymptotic range.

Asummaryofthegrid-dependentuncertaintiesisgivenin Ta-bles1and2forthecruiseandhoverconditionrespectively. Note thattheairframeiscomposed ofthefuselage,box-wingsand na-celles. In the cruise and hover conditions the uncertainties esti-matedforthetime-averagedpropeller performancequantitiesare relativelysmall.Fortheairframeliftanddragsomewhatlarger un-certainties are estimated. In the cruise condition, this is mainly duetoflow separationfromtheexhaustsontherearofthe fuse-lage,since noactiveflow was simulatedthroughtheexhausts.In the hover condition, the largely separatedflow from the leading andtrailingedges ofthe wings,dueto thevery largeinflow an-glebythemainrotordownwash,isthemaincauseofuncertainty. ThisflowseparationislaterdiscussedinSection3.2.Notethatthe largeGCIuncertaintyfortheairframedraginthehovercondition is aresult ofthe very low apparent order of 1

.

0, which maybe causedby the coarsegrid beingoutside ofthe asymptoticrange. The Richardson extrapolation uncertainty estimate E1 is signifi-cantlylower for thisquantity. Overall, asthe focus ofthis study is on the propeller performance, the uncertainties were deemed acceptable.

3. Results

Sincetheinteractions aresignificantlydifferent, theresultsfor thecruiseconditionandhoverconditionare treatedseparately in thenexttwosections.

3.1. Cruisecondition

Inthecruisecondition,thepropellersexperiencetherotational flowfield at the wingtip and are close to the main rotor slip-stream. In this condition, the main rotor and propellers operate atanadvanceratio(definedasV_∞

/(

ω

R

)

with

ω

inrad

/

s)above 0

.

5.InFig.4thissituationissketchedschematicallyandinFig.5

theflowfieldthrough the left propelleris visualisedby means of streamtraceswithvelocitycontours.Therespectiveinfluenceofthe wingsandmain rotor onthe propeller performance is treatedin thissection. Aswell, theupstream influence ofthepropellerson the wings is discussed. Mainly results for the left propeller are

Fig. 4. Sketch ofmain rotor,box-wing andpropeller aerodynamicinteractionin cruise,indicatingthedirectionofthemainrotorthrustTmrandpropellerthrust Tp.

Fig. 5. For thePNWFRconfigurationincruise,theflowfieldisvisualisedthrough theleftpropeller,with velocitycontoursonstreamtracesandonaplane0.18Rp upstreamofthepropeller.Onlytheleftbox-wing,nacelleandpropellerareshown, markedwithcontoursofthepressurecoefficientCp= (p−p∞) /q∞.

Table 3

Time-averagedleftpropeller per-formanceforthePNRandPNWFR configurations relativetothe PN configurationincruise.

PNR PNWFR T +2.9% +0.7% P +2.2% −7.8% ηp +1.0% +7.0%

given, since the results in this flight condition are qualitatively similar forboth propellers.All resultsare atzero angleofattack andsideslipunlessstatedotherwise.

In Table3the left propellerperformance is givenforthePNR andPNWFR configurationrelativeto theperformance atzero an-gleofattackforthePNconfiguration.ForthePNconfiguration,the bladepitchwastrimmedtoaspecificcruisethrusttarget.Forthe PNRconfiguration,thisbladepitchanglewasmaintainedtosee di-rectlytheeffectofthemainrotorslipstreamonthepropeller per-formance. However, forthe PNWFR configurationthe blade pitch angle was trimmed by a seriesof simulations ata reduced grid, in order to arrive ata thrust close to the initial target. The rea-son for trimming is that the interaction withthe wingincreases the propellerthrust drastically,aswillbeexplainedlater. The ro-tational speed andfreestream velocity were maintained for both configurations.FromthePNRconfigurationresults,conclusionscan be drawnontheeffectofthemainrotor onthepropeller perfor-mance.InthisconfigurationthepropellerthrustT andshaftpower

Fig. 6. Thebladesectionthrustrelativetothetime-averagedbladethrustisplotted foreachcircumferentialpositionoftheleftpropellerincruise.Theeffectofthe wingsisshownbymeansofcomparisonofthePNRandPNWFRconfiguration.

Fig. 7. Time-averagedaxialvelocitycomponentisplottedontheplanevisualisedin Fig.5,upstreamoftheleftpropellerincruise.Theeffectofthepropellerisshown bymeansofcomparisonoftheNWFRandPNWFRconfiguration.AnnotationA.and B.indicatetheeffectoftheupperandlowerwingflapedgegap.

P isincreasedasthepropellerexperiencesanon-zeroangleof at-tackfromthepotentialeffectofthemainrotoranditsslipstream. However,notethat nomainrotorslipstreamimpingementonthe propeller disk occurs. The propulsive efficiency

η

p

=

T V∞

/

P is slightlyincreasedaswellbythesamemechanism.Thismainrotor effect is slightly reduced for the right propeller, since it experi-ences a reduced downwash on the retreating blade side of this clockwiseturningmainrotor,confirming thefindingsbyWentrup et al. [8].

ForthePNWFR configuration,areduction inbladepitchangle was required to approximately maintain the thrust target as set forthePNconfiguration.The propulsiveefficiencyis significantly increasedby 7% as a resultof installation.As it was shown that forthePNRconfigurationthepropulsiveefficiencywasslightly in-creasedaswell,partofthis7% increaseisaresultoftheangleof attackinducedbythemainrotor.

Toinvestigate wherethe differences inpropeller performance forthe PNRandPNWFRconfigurationoriginate, inFig.6contour plotsofthe bladesection thrust areshownforthe leftpropeller. Thewhiteoutlineoftheupstreambox-wingenablescorrelationof thethrust variationwiththewinglocation.InthePNR configura-tionthe propeller blades experience a sinusoidal thrust variation in time, which is typical for a propeller at angle of attack. The maximumbladethrust isexperienced whenthe blademovesup, becausethen the angle of attack effectof the main rotor

down-Fig. 8. ThebladesectionefficiencyTV∞/Pisplottedforeachcircumferential po-sitionoftheleftpropellerincruise.Theeffectofthewingsisshownbymeansof comparisonofthePNRandPNWFRconfiguration.

Fig. 9. Time-averagedtangentialvelocitycomponentisplottedontheplane visu-alisedinFig.5,upstreamoftheleftpropellerincruise.Theeffectofthepropeller isshownbymeansofcomparisonoftheNWFRandPNWFRconfiguration.

washresultsinthelargestbladesectionangleofattack.This con-firmstheobservationbyWentrupet al. [8] forthesamecompound helicopter configuration.Forthe PNWFR configuration, additional non-uniformitiesare present. Amaximuminbladesection thrust isfoundaroundbladephaseangle

ϕ

=

180 degasdefinedinFig.1. Thismaximumoccurswhenthepropellerbladeisdownstreamof theupper wing.A suddenincrease inthrust isalsonoticeable in theproximityofthelowerwing.Botheffectsareexplainedbelow. The axial velocity component Va upstream ofthe propeller is plotted in Fig. 7 asdefined in Fig. 1. It is plottedon the plane thatwasvisualisedinFig.5.ContourplotsfortheNWFRand PN-WFR configuration are given, so without and with the propeller present. At the location of the peaks in thrust due to the box-wing in Fig. 6,a clear velocity deficit of the wake of the wings can be distinguished inthe axial velocity field.Especially notice-ablearethedeficitsatthelocationoftheflapedgegapsindicated by annotationsA.andB.The axialvelocity deficitresultsina lo-cal advance ratio reduction and thus thrust increase. Comparing thecontourplotsfortheNWFRandPNWFRconfiguration,the up-stream accelerationof the flowinduced by thepropeller loading isclearlyvisible,resultinginstrongflowaccelerationbetweenthe wingsasannotatedbylabelBinFig.5.

Pusher-propellers mounted at the tip of planar wings are as-sociated withbeneficial interaction effectsin termsof propulsive efficiency, asisknown fromi.e.Refs. [9–12]. Toshow where the

Fig. 10. Forthreeconfigurations,PNWFR,PNandNWFR,theradiallyand circumfer-entiallyaveragedaxialandtangentialvelocitydistributionthroughtheleftpropeller approximatestreamtubeisplottedincruise.Thestreamtubewasapproximatedby acylinderofradiusequaltothepropellerradiusfrom3radiiupstreamto3radii downstreamofthepropeller.

propulsiveefficiencyincreaseiscomingfrombetweenthePNRand PNWFR configuration,in Fig.8acomparisonofthe bladesection efficiencyover the propeller disk is given. This is definedasthe ratioofblade section contributionto thrust T andblade section contribution to power P, multiplied by the freestream velocity magnitude V_∞. Although over the entire disk the blade section efficiencyisraised forthePNWFR configurationcompared tothe PNRconfiguration,especiallyonthe inboardside ofthepropeller diskandforbladesectionsatlowradiialargeincreasecanbe no-ticed,asexplainedbelow.

The distribution of efficiency increase corresponds directly to the negative tangential velocity field shown in Fig. 9,as defined inFig. 1with thetangential component positive inthe direction of propeller rotation. This negative tangential velocity is result-ingfromthepressuredifferencebetweenthepressuresideofthe lowerwingandsuctionsideoftheupperwing.Thenegative tan-gential velocity component increases the propeller blade section angleof attack and tilts thesection liftmore inthe direction of thethrustandlessinthetangentialdirection,reducingthe contri-butiontothetorqueorshaft power.Comparingthecontourplots forthe NWFRandPNWFR configuration,a small upstreameffect of the propeller can be noticed by an increase of the negative tangentialvelocity.Sinceapropellerdoesnotinduceanupstream tangentialvelocityfield,thischangeintangential velocityislikely aresultfromachangeinwingloadingandits inducedtangential velocityfield.

Toqualitativelyshowhowtheaerodynamicinteractionschange theflowfieldthroughthepropeller,inFig.10velocitycomponents through the approximated streamtube of the left propeller are plottedforthePNWFR, PNandNWFRconfiguration.The stream-tube is approximated by a cylinder with a radius equal to the propellerradiusandquantitiesaretime- andspace-averagedover disksfrom3Rp upstreamto3Rpdownstreamofthepropeller.For

thePNWFRandNWFRconfiguration,alargepeakinaxialvelocity

Fig. 11. Time-averagedwingloadingcomparisonbetweenPNWFRandNWFR con-figurationfortheleftwingincruise,withcoordinateypasdefinedinFig.1andc thewingchord.Shadedareaindicatestime-variation.AnnotationA.andB.indicate theeffectoftheupperandlowerwingflapedgegap.

canbe noticedupstreamofthepropeller,whichisaresultofthe acceleration of the flow around the wings,since the streamtube intersectspartofthewings.Thisaccelerationisalsovisibleinthe velocitycontoursonthestreamtracesinFig.5andinthereduced pressurecoefficient onthewingandnacellesurface, especiallyin betweentheupperandlower wingasindicatedby annotationB. The axial velocity increase isslightly higherforthe PNWFR con-figuration because ofthe induced axial velocity by the propeller. Comparing the PNWFR and PN configuration, the axial velocity componentjustupstream ofthepropeller isloweronaveragefor theinstalledcase,likelyduetothewakeofthewingsandchange inflowdirectiontomoretangentialflow.Thedifferencesupstream ofthe propeller impactthe slipstreamofthepropeller:The axial velocitycomponentlevelsofftoaslightlylowervalue.

Looking at the tangential velocity plot, a growth in negative tangentialvelocitycomponenttowardsthepropellerispresentfor the PNWFR andNWFRconfiguration,induced bythe wing. Com-paringthePNWFRandPNconfigurationresultsdownstreamofthe propeller,thetangentialvelocitycomponentisconsiderablylower forthe installedcase. ComparingtheNWFRandPNWFR configu-rationresults,thetangentialvelocityfieldaroundthewingtiphas changed directionduetothe propeller.Asthe tangential velocity magnitudeislower,thetangentialkineticenergyintheslipstream hasbeenreducedbytheinteraction.

FromFigs.7through10itisclearthatpartofthewing experi-encesadifferentflowfieldasaresultofthepropellers,inparticular inaxialvelocity.Therefore,itcanbeexpectedthatthewing load-ingmayalsochange.Asthereisafinitenumberofpropellerblades andthepropellersrotate,theyalsointroducetime-dependent vari-ationsinwingloading.Toidentifychangesinwingliftanddragby thepropellers,inFig.11thetime-averagedspanwisewingloading isplottedfortheleftwinghalf.Notethattherightwinghalf expe-riences higherwingliftduetothereducedmainrotordownwash onthatside[8],buttheinteractionphenomenaaresimilar.Results fortheNWFRandPNWFRconfigurationsareshown,includingthe

Fig. 12. LeftpropellerthrustT andpropulsiveefficiencyηp= (T V∞cosα) /P are

plottedagainstangleofattackα atconstantbladepitchanglefor V∞=220kt (113.2m/s).AcomparisonisshownbetweenthePNWFandPNconfigurationand quantitiesarewithrespecttoPNconfigurationquantitiesatzeroangleofattack.

time-dependentvariationindicated bytheshaded areas.The pro-pellersintroducelocally aslightreduction inupperwingliftand larger increase in lower wing lift, as well as a time-dependent variation.Wingdrag is increasedlocallyfortheupper andlower wing,resultingin anetreduction oflift-over-dragratio.At span-wiselocationsawayfromthepropeller,theeffectofthepropeller reduces.

Because of the strong aerodynamic interaction between the propellersand the box-wings, an investigation of the interaction foroff-designconditionsisessentialtounderstandtheflight per-formance and dynamics of this vehicle. The effects of angle of attack andsideslip were studied withthe PNWF and PN config-urations,leavingthemainrotoreffectoutforthisstudy.InFig.12

the left propeller performance is plottedas function of angle of attack.Resultsaremaderelativewiththeperformanceatzero an-gle of attack for the PN configuration. Forthe PNconfiguration, thethrustandpropulsiveefficiencyincreasewithangleofattack, symmetricallyaroundzeroangleofattack.ForthePNWF configu-ration,thewingintroducesan increaseinthrustandefficiencyat zeroangleofattackdueto thetangentialvelocity field andwing wakes. Since the strength of the tangential velocity field is cor-relatedto wing lift andwing lift increaseswith angle of attack, anincreaseinangleofattackincreasesthrustandefficiencymore thanfortheisolatedpropeller.Thesymmetryaroundzeroangleof attackisthereforealsolostasaresultoftheaerodynamic interac-tionwiththewing.

Toshowtheeffectsofsideslip,therightpropeller wasstudied witha positive sideslip angle to avoid significant fuselage inter-action. In Fig. 13 the right propeller performance as function of sideslipisplottedforareducedmanoeuvringcruisespeed.Again, atzerosideslipthethrustandefficiencyarehigherforthePNWF configurationcomparedto thePNconfigurationdueto theeffect ofthe wing. The difference inthrust andefficiency betweenthe PNWFandPNconfigurationresultgraduallyreduceswith increas-ing sideslip as the wing lift reduces and thus the effect of the wingonthepropeller reducesforthePNWFresult. Notethatthe propelleris operatingatadifferentadvanceratio,such thatwith increasing angleof attack the efficiency forthe PNconfiguration reducesalreadyforsmallangles,oppositetotheresultsofFig.12.

3.2.Hovercondition

Themainrotorplaysonlyaminorrole intheaerodynamic in-teractionsexperienced by the propellers in the cruise condition.

Fig. 13. RightpropellerthrustT andpropulsiveefficiencyηp= (T V∞cosβ) /P are

plottedagainstsideslipβatconstantbladepitchangleforareducedcruisespeed formanoeuvringofV∞=140kt (72.0m/s).Acomparisonisshownbetweenthe PNWFand PNconfigurationandquantitiesarewith respecttoPN configuration quantitiesatzerosideslip.

Fig. 14. Sketchofmainrotorandpropelleraerodynamicinteractioninhover, indi-catingthedirectionofthemainrotorthrustTmrandpropellerthrustTp.

However, the lackof shielding ofthe propellersallows fora sig-nificantmainrotor interactionatlowairspeedandparticularlyin thehoverconditionwherethemainrotordownwashimpinges on the propellers, as sketched in Fig. 14. The flowfield through the leftandrightpropellerisvisualisedinFig.15forthePNWFR con-figuration by streamtraceswith velocity contours and thisfigure indicates the large angleofattack experienced by the propellers. ThevelocityismadedimensionlesswithVd,thetheoreticfarfield downwash velocity of the main rotor based on momentum the-ory Vd

=



_2T mr

ρ∞πR2 mr

[29].Notethat,inordertocounterthetorque of the main rotor, the right propeller blade pitch angle is setto produceareversethrustinthehovercondition.Notonlythe pro-pellers butalso thewingsexperience thevery largenegative an-gle ofattack duetothe mainrotor downwash,resulting ina net download, aswas alsoshown by Lynn [14]. This resulted in the choiceofabox-wing designfortheAirbusRACER,inorderto re-ducetheoverallsurfaceaffectedbythedownwash.Itcanbeseen that thewings introduceadditional disturbancesto theinflow of especiallytheleftpropellerwithunknownimpactonthepropeller loading.Since theright propelleris producingreverse thrust, the wingsareinitsslipstream,andtheeffectofthewingsonthis pro-pellerarelikelydifferent. Becauseoftheseunknowns,theimpact of this complex interactional flow on the propeller performance andunsteadyaerodynamicloadingwasinvestigated.

Tofurther illustrate theflowfieldexperienced by the left pro-peller,inFig.16thevelocitymagnitudeisplottedwithstreamlines on a plane throughthe propeller forthe NWFRconfiguration, so

Fig. 15. ForthePNWFRconfigurationinhover,theflowthroughthepropellersisvisualised,withvelocitycontoursonstreamtracesandonaplane0.18Rpupstreamof theleftpropeller.Thefuselageisnotshown.AnnotationA.indicatesthewakeofthewingsandnacelle,B.mainrotordownwashdeflectedbyupperwingandC.thearea undisturbedbythewings.

Fig. 16. Thevelocityfieldinaplanethroughtheleftpropellerinhoverisshownfor theNWFRconfiguration,sowithoutpropellerpresent.

without propellers present. The propeller will experience a non-uniformdistributionofdownwashbythemainrotor.Furthermore, anareaofreducedvelocityonthelowersideofthenacellecanbe distinguished. Thiswake isformed because thenacelle is shield-ingthisareafromthemainrotordownwash.Consideringthatthe wakeismorepronounced onthe inboardsideofthenacelle,the shielding of the main rotor downwash by the wings likely also playsaroleinthedevelopmentofthiswake.Theflowfieldinthis plane at the location of the rightpropeller is very similar, since the main rotor loading is equal on both sides inthe hover con-dition. The main difference is the out-of-plane swirl component introduced by the main rotor, which cannot be observed in this figure.

To establish a quantitative indication of the aerodynamic in-teraction effects,simulations were performedwiththe propeller-nacelle-main rotor configuration PNR and

nacelle-wing-fuselage-Fig. 17. Bladethrustisplottedrelativetothe time-averagedbladethrustduring afullrotationinhoverfor thePNRand PNWFRconfiguration.Phaseangleϕis definedinFig.1andthequantitiesareplotteddividedthroughthePNWFR time-averaged value.AnnotationA.indicatestheeffectofthewakeofthe wingsand nacelleandB.theeffectofmainrotordownwashdeflectedbytheupperwing.

main rotor configurationNWFR andwere compared toresults of the complete configuration PNWFR. Differences in integral pro-pellerloadingwerefound: Althoughthebladepitchangleandall other operating conditions were kept constant between the PNR andPNWFR configuration,for theleft propeller thethrust in the PNWFRconfigurationwasfoundtobe10

.

5% higher.Thisdifference isthereforelikelyaresultoftherole thewingsplayinthe inter-action. The thrust-over-power ratio T

/

P forthe left propeller in thePNWFRconfigurationwasslightlyreduced,althoughthiscould be directlyaresultoftheincreasedthrust.Contrary,fortheright propeller the thrust was 1

.

5% reduced in the PNWFR configura-tion andthe thrust-over-power ratiodid not significantly change betweenthetwoconfigurations.

In order to investigate the differences between the propeller bladeloadingforthePNRandPNWFRconfiguration,inFig.17the blade thrust evolutionovera completerotation isplottedforthe

Fig. 18. Thebladesectionthrustrelativetothetime-averagedbladethrustisplotted foreachcircumferentialpositionoftheleftpropellerinhover.Theeffectofthe wingsisshownbymeansofcomparisonofthePNRandPNWFRconfigurationat equalbladepitchangle.

left andright propeller,both relative to the time-averagedblade loadingfor the PNWFR configuration. Although not quantified in thefigure,thepeak-to-peakvariationinbladethrustisvery signif-icant.Forthe left propeller thepeak-to-peak variationis approx-imately equal to 60% of the maximum required thrust in hover forcounter-torque (leftpropeller 100% thrust, rightpropeller 0% thrust),butnotethatthisplottedcaseisnotthemaximumthrust requiringcase. In general, theloading evolution isapproximately sinusoidal for both propellers in both configurations. The phase and amplitudes of the PNR and PNWFR result are very similar forbothpropellers.Note that althoughforthe PNRconfiguration nowingsandfuselageare present,theblade loadingevolutionis notperfectlysinusoidalforeitherpropeller,whichisnormally ex-pectedforapropelleratangleofattack.Thisislikelyrelatedtothe non-uniformityinthemainrotordownwashthatwasillustratedin Fig.16.Whenthewingsarepresent,fortheleftpropellera signif-icantlyincreasedthrust isnoticeable inbetween

ϕ

=

90 deg and 160 deg,indicated byannotationA.,afterwhicha relative reduc-tionofthrustoccursindicatedbyannotationB.

For the right propeller the differences between the PNR and PNWFRconfigurationresultsareverysmall.Thisisthoughttobe, becausethewingsareintheslipstreamofthepropelleranddonot significantlydisturb theinflowto thepropeller.However, asmall thrust decreaseis noticeablebetween

ϕ

=

130 deg and270 deg, whentheblademovesbehindandabovethewings.Asmalllocal angleofattack reductioninducedbythewingscouldhavecaused thisthrustreduction.

For the left propeller the differences between the PNR and PNWFR configuration are larger than for the right propeller and therefore require further investigation. Fig. 18 allows correlation of the thrust variation with the wing location for the left pro-peller through contour plots of the blade section thrust for the PNRandPNWFRconfiguration.Thisfigureshouldbeviewed along-side Figs. 19and 20, which show a comparisonof the axialand tangentialvelocitydistributionfortheNWFRandPNWFR configu-ration,sowithoutandwithpropeller respectively.Thesecontours areplottedontheplanevisualisedinFig.15(right)justupstream oftheleftpropeller,whenseenfrombehind.Notethatinthis fig-urealso a frontview sketch ismadeof thepropeller disk anda divisionismadeinthreezonesannotatedA.,B.andC.An explana-tionofthepropellerloadingforeachzoneisgiveninthefollowing paragraphs.

Inzone A.thepropellerinthePNWFR configurationdrawsair fromthewakeofthenacelleandwingsthatwas madevisiblein Fig.16.InFig.19indeedan almostzeroaxialvelocitycomponent

Fig. 19. Time-averagedaxialvelocitycomponentisplottedontheplanevisualised inFig.15,upstreamoftheleftpropellerinhover.Theeffectofthe propelleris shownbymeansofcomparisonoftheNWFRandPNWFRconfiguration.

Fig. 20. Time-averagedtangentialvelocitycomponentisplottedcorrespondingto Fig.19.

canbe observedinthisarea whenthepropellerisnotpresentin theNWFRconfiguration.Thealmoststaticinflowconditionresults inalargeangleofattackforthepropellerbladesectionsandthus highthrust,explainingthesuddenthrustincreasefrom

φ

=

90 deg to 160 degcompared tothe PNRconfiguration.InFig. 19apeak inaxialvelocityforthePNWFRconfigurationisvisibleinthisarea, induced bythe propellerasa resultofthe highthrust.Notethat alsoa largenegative tangentialvelocity componentisinducedby thepropeller inthisarea,showninFig.20.Typicalofapropeller operatingatstaticconditionisthatairisdrawntoitfromitswider surroundings and this is also clearly visible in Fig. 15, where a spanwise flow is visible in between the wings from inboard to-wardsthepropeller atits outboardlocation.Theformationofthe negative tangential velocitycomponentupstream ofthepropeller maybe aresultof thelower wingguidingthepropeller induced flowinthattangentialdirection.

In zoneB. an area ofpositive axial velocity canbe noticed in theNWFRresultofFig.19,whichisaresultofthedeflectionofthe main rotor downwash by the upper wing towards the propeller. ThisdeflectioncanalsobeobservedbythestreamtracesinFig.15. Thiseffectoftheupperwingontheaxialvelocityisnoticeablein thethrustdistributionplotinFig.18byalocalreductioninthrust, as alsopreviously identified in the blade loading evolution from

φ

=

160 degonwards.Theaxialvelocity inflowreducestheangle ofattack ofthe propeller bladesectionsandthus reducesthrust. Therefore,in thePNWFR resultofFig.19,although an additional

increaseinaxialvelocitybythepropellerisseeninthisarea,itis oflowermagnitudethaninzoneA.

In zone C. no large differencesbetween the PNWFR andPNR configurationin terms ofpropeller thrust are noticeable. The in-flow to the propeller is free of objects for both configurations and operates in a way typical of a propeller at angle of attack. The increasedintegratedpropeller thrust of10

.

5% inthe PNWFR configuration, which was mentioned earlier, should therefore be explainedby the effects inzone A. andB.Apparently, the thrust increasing effectofthewake isstrongerthan thethrust decreas-ing effect by the deflection of the main rotor downwash by the upperwing,resultinginanetthrustincrease duetothepresence ofthewings.

4. Conclusions

As part of the Clean Sky 2 (CS2) research programme, aero-dynamic interactions occurringfor the AirbusRACER (Rapid And Cost-EffectiveRotorcraft) were investigated ina cruise andhover condition. Abreakdownwas presented oftheaerodynamic inter-actionsbetweenthelateralrotorsorpropellers,thebox-wingsand the main rotor, focussing mainly at the effects on the propeller loading.

Since the propellers ofthis compound helicopter are situated belowthe main rotor and themain rotor slipstream passesover propellersin the cruise condition, the effectwhich the main ro-tor has on the propellers is limited to an angleof attack effect. A smallnegative angle of attack is induced to the inflow of the propellers,resultinginasinusoidallyvaryingpropellerblade load-ing.Contrary,inthehoverconditionthedirectimpingementofthe main rotor downwash on thepropellers resultsin a very signifi-cantsinusoidallyvaryingbladeloadingduetotheinflowatalmost perpendicularangletothepropelleraxis.Anysuchcompound he-licopterlayoutwhereapropellerissituatedbelowthemainrotor would experience a similar interaction, if no other aerodynamic surfaceswouldbeshieldingit.

Since the propellers of this compound helicopter are located behind thewingat thewingtip,they ingest therotational veloc-ityfieldformedbythepressuredifferencebetweentheupperand lowerwingsurfaces.Consideringthattherotationdirectionofthe propellersisagainst the rotationofthe flow aroundthe wingtip, anincreaseinthrustoccurs.Furthermore,thelowaxialvelocityof thewingwakeisingested,alsoresultinginalocalpropellerthrust increase.Comparisonwiththeisolatedpropeller performance has ledto the conclusionthat installation increasespropeller propul-siveefficiencyby7% ifthebladepitchisreducedtomaintainequal thrust.Themajorityofthisgaincanlikelybeascribedtothe rota-tionalflowfield. Duetothecloseproximityofthepropeller tothe wing, an upstream effect onthe wingcan however be expected. Forthishelicopter,thepropellersincrease winglift,througha lift increaseofthelowerwingandasmallerliftdecreaseoftheupper wing. Thepropellersincreasedrag oftheupperandlower wings, therebydecreasingthewinglift-over-dragratio.

This close couplingof propeller andwing loading has conse-quencesfor incidence angleeffects, forinstance during manoeu-vres.Forthishelicopter,there isastrong positivecorrelation be-tween wing liftand propeller thrust andefficiency inthe cruise condition. This significantly changes angle of attack and sideslip angleeffectscomparedtothoseofanisolatedpropeller.

SinceinthehoverconditionthepropellersontheAirbusRACER are used tocounter thetorque of the mainrotor, an asymmetry arisesbetweentheleft andrightpropeller.Forthe left propeller, the wingintroduces disturbances to theinflow. This causes local changes inthe sinusoidalpropeller blade loadinginduced by the mainrotor.Theleftupperwingdeflectsthemainrotordownwash towardstheleftpropeller,resultinginanaxialvelocitycomponent

whichdecreaseslocallythethrust.Thewakeformedbythenacelle andwingsinthemainrotordownwashcausestheleftpropellerto locallyoperatesimilartoinstaticcondition,drawinglow momen-tum flow from the wake. This increases the thrust. Overall, the wingsincrease theleftpropeller thrustby 10

.

5% withslightly re-ducedthrust-over-powerratioforequalbladepitch.Fortheright propeller,thewingsare inits slipstream.Theyleadto adecrease ofrightpropellerreversethrustby1

.

5% withconstant thrust-over-powerratioforequalbladepitch.

Although thepresented resultsare very specific tothe Airbus RACER geometry,in generalverysignificant aerodynamic interac-tion effects can be expected when a main rotor, propellers and wingareinproximitytoeachother.Abreakdownofaerodynamic interaction effectsby leavingaway partsofthe geometry,as pre-sentedhere,canprovidethenecessaryinsightintotheseeffects. Declarationofcompetinginterest

Theauthorsdeclarethattheyhavenoknowncompeting finan-cialinterestsorpersonalrelationshipsthatcouldhaveappearedto influencetheworkreportedinthispaper.

References

[1] H. Yeo, Design and aeromechanics investigation of compound helicopters, AerospaceScienceandTechnology88(2019)158–173,

https://doi

.org/10.1016/ j.ast.2019.03.010.

[2] H.Yeo,W.Johnson,Optimumdesignofacompoundhelicopter,Journalof Air-craft46(2009)1210–1221,

https://doi

.org/10.2514/1.40101.

[3] J.Enconniere,J.Ortiz-Carretero,V.Pachidis,Missionperformanceanalysisof aconceptualcoaxialrotorcraftforairtaxiapplications,AerospaceScienceand Technology69(2017)1–14,

https://doi

.org/10.1016/j.ast.2017.06.015. [4] Airbus, Airbus helicopters reveals racer high-speed demonstrator

config-uration, https://www.airbus.com/newsroom/press-releases/en/2017/06/Airbus -Helicopters-reveals-Racer-high-speed-demonstrator-configuration.html,2017. [5]C.Lienard,I.SalahelDin,T.Renaud,R.Fukari,RACERhigh-speed

demonstra-tor:rotorandrotor-headwakeinteractionswithtailunit,in:AHSInternational 74thAnnualForum&TechnologyDisplay,2018.

[6] S.A. Andrews, R.E. Perez, Comparison of box-wing and conventional air-craftmissionperformanceusingmultidisciplinaryanalysisandoptimization, AerospaceScienceandTechnology79(2018)336–351,

https://doi

.org/10.1016/ j.ast.2018.05.060.

[7]T.C.A.Stokkermans,M.Voskuijl,L.L.M.Veldhuis,B.Soemarwoto,R.Fukari,P. Eglin,Aerodynamicinstallationeffectsoflateralrotorsonanovelcompound helicopterconfiguration,in:AHSInternational74thAnnualForum& Technol-ogyDisplay,2018.

[8]M.Wentrup,J.Yin,P.Kunze,T.Streit,J.-H.Wendisch,T.Schwarz,J.-P.Pinacho, K.Kicker,R.Fukari,AnoverviewofDLRcompoundrotorcraftaerodynamicsand aeroacousticsactivitieswithintheCleanSky2NACORproject,in:AHS Interna-tional74thAnnualForum&TechnologyDisplay,2018.

[9] J.C. Patterson Jr, G.R. Bartlett, Effect of a wing-tip mounted pusher tur-boprop on the aerodynamic characteristics of a semi-span wing, in: AIAA/SAE/ASME/ASEE21stJointPropulsionConference,AIAAPaper1985-1286, 1985,

https://doi

.org/10.2514/6.1985-1286.

[10]J.C.PattersonJr,G.R.Bartlett,EvaluationofInstalledPerformanceofa Wing-Tip-MountedPusherTurboproponaSemispan Wing,TechnicalPaper2739, NASA,Hampton,Virginia,1987.

[11] J.M.Janus,A.Chatterjee,C.Cave,Computationalanalysisofawingtip-mounted pusherturboprop, JournalofAircraft33(1996) 441–444,https://doi.org/10. 2514/3.46959.

[12] L.R.Miranda,J.E.Brennan,Aerodynamiceffectsofwingtip-mountedpropellers andturbines,in:4thAppliedAerodynamicsConference,AIAAPaper1986-1802, 1986,

https://doi

.org/10.2514/6.1986-1802.

[13]L.J.Bain,A.J.Landgrebe,InvestigationofCompoundHelicopterAerodynamic InterferenceEffects,TechnicalReport67-44,USAAVLABS,1967.

[14] R.R. Lynn, Wing-rotor interactions, Journal of Aircraft 3 (1966) 326–332,

https://doi.org/10.2514/3.43742.

[15] A. Batrakov, A.Kusyumov,S. Mikhailov,G. Barakos, Aerodynamic optimiza-tionofhelicopterrearfuselage,AerospaceScienceandTechnology77(2018) 704–712,

https://doi

.org/10.1016/j.ast.2018.03.046.

[16]I.SalahelDin,C.Lienard,R.Huot,R.Fukari,RACERhigh-speeddemonstrator: tailunitverticalfinaerodynamicdesign,in:AHS International74th Annual Forum&TechnologyDisplay,2018.

[17]C.Öhrle,U.Schäferlein,M.Keßler,E.Krämer,Higher-ordersimulationsofa compoundhelicopterusingadaptivemeshrefinement,in:AHS International 74thAnnualForum&TechnologyDisplay,2018.

[18] J.Thiemeier,C.Öhrle,F.Frey,M.Keßler,E.Krämer,Aerodynamicsandflight mechanicsanalysisofairbushelicopters’compoundhelicopterRACERinhover undercrosswindconditions,CEASAeronauticalJournal(2019),

https://doi

.org/ 10.1007/s13272-019-00392-3.

[19]J.W.Boerstoel,A.Kassies,J.C.Kok,S.P.Spekreijse,ENFLOW,aFull-Functionality SystemofCFD Codes for IndustrialEuler/Navier-Stokes FlowComputations, TechnicalPublicationTP96286U,NLR,1996.

[20]O.J. Boelens,H. Ven, J.C. Kok,B.B. Prananta,RotorcraftSimulations Usinga Sliding-GridApproach,TechnicalPublicationNLR-TP-2008-779,NLR,2008.

[21]M.Laban, J.C. Kok,B.B.Prananta, Numericaltools forcontra-rotating open-rotor performance, noise and vibration assessment, in: 27th International Congress oftheAeronautical Sciences (ICAS2010), Nice,France,September 2010,pp. 19–24.

[22] A.F.Antoniadis,D.Drikakis,B.Zhong,G.Barakos,R.Steijl,M.Biava,L.Vigevano, A.Brocklehurst,O.Boelens,M.Dietz,M.Embacher,W.Khier, Assessmentof CFD methodsagainst experimentalflowmeasurementsfor helicopter flows, AerospaceScienceandTechnology19(2012)86–100,

https://doi

.org/10.1016/j. ast.2011.09.003.

[23] E.W.M.Roosenboom,A.Stürmer,A.Schröder,Advancedexperimentaland nu-mericalvalidationandanalysisofpropellerslipstreamflows,JournalofAircraft 47(2010)284–291,

https://doi

.org/10.2514/1.45961.

[24] B.Ortun,R.Boisard,I.Gonzalez-Martino,In-planeairloadsofapropellerwith inflowangle:predictionvs.experiment,in:30thAIAAAppliedAerodynamics Conference, AIAAPaper2012-2778,2012,

https://doi

.org/10.2514/6.2012-2778. [25] S.Wallin,A.V.Johansson,AnexplicitalgebraicReynoldsstressmodelfor in-compressibleandcompressibleturbulentflows,JournalofFluidMechanics403 (2000)89–132,

https://doi

.org/10.1017/S0022112099007004.

[26] F.R. Menter, Two-equation eddy-viscosity turbulence models for engineer-ingapplications,AIAAjournal32(1994)1598–1605,

https://doi

.org/10.2514/ 3.12149.

[27] I.B.Celik,U.Ghia,P.J.Roache,C.J.Freitas,H.Coleman,P.E.Raad,Procedurefor estimationandreportingofuncertaintyduetodiscretizationinCFD applica-tions,JournalofFluidsEngineering130(2008)078001,

https://doi

.org/10.1115/ 1.2960953.

[28] P.J.Roache,Quantificationofuncertaintyincomputationalfluiddynamics, An-nual ReviewofFluidMechanics29(1997)123–160,

https://doi

.org/10.1146/ annurev.fluid.29.1.123.