RECENT WIND TUNNEL TESTING EXPERIENCE OF CONTRA-ROTATING PROPELLERS 1RS-90-4.1.1 AE Warris, PM Render, 0 M Pozntak, ME Wood ‘Aircraft Re In recent years the ARA has been engaged tn a substantial” series of contra-rotating propeller (CRD) wind-tunnel based research programmes. All programmes have been undertsken within a UK club arrangenent Involving British Aerospace, Rolls- Royce, Dovty Aerospace and the ARA, with Department of Trade and. Industry and Royal Aerospace Establishment support. Advances in experimental techniques have led to the capability to perfora research In the following areas 2) Small scale isolated and installed CRP testing using conpressed alr driven turbine motors to study rotors of typically 0,4 m dlaneter; b) Large scale Isolated testing using electric motor driven CRP models of typically 0.8 = Aianeter ©) CRP performance studies using rotating shaft ounted balances at large and small scale; 4) Noise field studies using an acoust (cally lined wind tunnel “in parallel with CRP performance 2) Flow field surveys supported by ‘club’ laser and convent ional measurements. The ‘paper sets cut to review all of the above techniques within the broad context of experinental research on advanced rotor designs BPF_ Blade passing frequency Cp:Corac Drag. coefficient ChiscH” Nozzle discharge coefficient lis, Life coefficient Cp-Crower Power coafficient = power/e.NDS CriCrymusr Propeller thrust coefficient = ‘thrust /eN*D* also nozzle thrust coefficient Torque coefficient = torque/pl2D* Propeller dlaneter Frequency Advance ratio = V/ND Free stream Mach number Propeller rotational speed, rotat lons per’ second Nozzle pressure ratio Free stream static pressure Free stream velocity o Incidence Blade. ang 4 Inerenent Increment in body drag due to slipstream Propeller efficiency Polar angle Free stream density ‘Subsoriots Pertaining to front rotor Overall (Front + rear) Pertaining to rear rotor Naximun Mintaum prom MN Copyright © 190 by HCAS und ALAA. All ahs rere. 92 iret Assoclation Ltd, Bedford Mal TPF, WC ABBREVIATIONS Able Able Corporation, Anaheim, Ca, USA aC Alternating current ADC Analog to digital converter ARA Aircraft Research Association Ltd BAe «British Aerospace (Commercial Aircraft) Led, Hatfield, England oe Centrifugal force CED Computational Fluid dynaates CRP Contra-rotat ing propellers a Dowty Aerospace Gloucester Ltd DIL Department of Trade & Industry High pressure (air) He Hertz, LTA Laser transit anesonetry Mod Ministry of Defence MST Mach Stmulat fon Tank MIR Microphone Traverse Rig Ne Nonerical control RAE Royal Aerospace Establisheent RR Rolls-Royce ple SPARV Source Patch & Ring Vortex SRP Single rotat lon propeller m Tech Development. Ine, Dayton, Ohie, USA TWE—ARA transonte wind tunne! Lrronvctios The objective of this paper Is to recount recent ARA experience related to contra-rotat ing. propeller (CRP) research. Whilst AFA involvement in the CRP design process will be reported elsewhere, the al here ie to illustrate the diverse nature of the large and small scale GRP experimental studies tn which ARA has participated Experimental work on advanced propellers began at ABA in the mid 1980s, with the aim of establishing capability to make accurate aerodynamic, acoustic and performance measurenents on open rotors in the ARA 'trangonic. wind tunnel! ‘The sequence of experinental programmes to date Is shown In Figure 1. "The dominant feature of all of the. progranne: outlined in Figure 1 is that they are jointly Funded ‘club’ activities; those club activities have systematically Involved the appropriate specialists. drawn” from industry including particularly Rolls-Royce, British Aerospace and Dowty Aerospace as well as significant work placed with smaller companies; all vork has received Substantial support from the UK Department of Trade and Industry and has benefited from active involvement of RAE specialists, As shown In Figure 1 the emphasis of the programe has moved In recent years from single rotating propellers (SRP) to cover the more challenging CRP Fesearch needs. In total these needs are overall and individual rotor _ performance (thrust, torque, power, effteteney) at on- and off-design conditions; )3.847, 9,02 uc cue roams 74 a Le (3/8) 86 SERA reemuearnan| 89 i. ee SiMe sie ey a aname Feces es | aw ime [LARGE SCALE 3/6) m8 ca TEE ts aut | ORE a a FIG 1 ARA INVOLVEMENTS IN UK OPEN ROTOR RESEARCH CLUB. PROCRAIMES b) CRP near field and far Field noise spectra at fon and off-design conditions: ©) Isolated CRP performance/noise data base for ‘evaluation of CFD predict{ons; 4) Flow field and blade/nacelle pressure data to persit evaluation of local performance Contributions and to enable” the. empirical elesents in prediction methods co be Improved or replaced; ¢) Blade strain gauge and blade ‘live? twist data to supplement’ rotor Finite elenent structural analyses and to refine aerodynamic ‘live! blade CFD eodet ling: 4 rotor slipstream surveys at an ald to jation and alr intake design activities: 8) Techniques of proven capability for CRP project developaents: hy Inst n Jed CRP performance data. AIL of the above objectives have played @ pare in the foraulation of the research studies discussed her In the following sections sa ore Fecent CRP activities tes yoted that many of the techniques employed in the SRP and CRP research studies derive, in part, from the established ponered testing techniques in use at AA” In’ support of fsolated and. installed Convent tonal turbofans sRP_AND cRP_1Ns Dies Figure 2 shows an overview of a large and small scale nodel development testing methodology Leading fo @ well engineered installation for an SRP or CRP powered aircraft. The five test phases A to E are used in a combined manner to assess overall rotor and installation aerodynamic performance in Tine with "an appropriate thrust/drag account ing Framework. ‘The large scale rig of phase F x used to extend understanding of tsolated roter performance as well ag to study (or further study) 93 rotor nofse and to examine alr intake/rotor Interactlon performance. Further discussion of large scale studies is deferred to section 4 below. ‘The five small scale test phases iNustrated tn Figure 2 yield an assessment of Installed SRP or CRP Interference drag, ‘This capability to evaluate those elements overall drag. which may be classified “interference’ 1s 3 vital factor in the develo of high speed rotor powered afreraft. It 1s cl that Figure 2 presente an over-rinplified version of the practical accounting steps involved. Factors not listed in the scheme Include the substantial pre-mode! design/prediction staj both rotor performance and. installation The assessment” of overall aircraft/rotor performance 1s the province of airframs and Propulsion specialists; recognising this, the Femainder of this paper concentrates on the research facilities which provide the data base for Such assessments SMALL SCALE EXPERIMENTS, eT ee Emin IN Ae TE, "Suramar, Soe Bsr gn meus SE sae xs nas MSR "8 pew oe, saw mrsee won, SE Caan cis cmt Fe on ow @ seiner of argmral sn wah cx Gl and aos ‘Shp fat pas Sade - LARGE SCALE EXPERIMENT ‘Prnse LARGE SOLE SHO oR CRP ROTR/ ANE FIG 2 OVERVIEW OF SRP OR CRP INSTALLATION DEVELOPMENT PROGRANNE3. SMALL SCALE CRP EXPERIENCE 3.1. Special Rig for Izolated CRP Tests In the period 1988/90 2 small scale model CRP technique thas been evolved to cater for tsolated and installed testing In answer to the cosbined requirenents of British Aerespace, Dowty Aerospace, the DTI and ARA ‘The thrust, torque and power account ing schemes for an SRP or GRP Invariably require a clear description of the contributions of the rotor, the nacelle and the nozzle elenents of the powerplant installation, Figure 3 shows the main features of a high speed wind tunnel rig developed specifically Tor support of CRP Isolated snall scale studies At the rig design stage, considerations of rig structural survival In the event of loss of Blade, or blades, led to a robust strut mounted ascenbly: a mon-netric shield is provided to sninimlse concerns regarding the effects of CRP slipstream on the downstream rig elenents. The rig {a mounted on a well established underfloor balance; the key feature of the rig te th provision of two shaft mounted rotor force and smonent balances; these balances are used for evaluation of roter/hub thrust, torque and power (Gee section 3.2) ae 1 sie emmome b nse ex vole | ae Lt = | Fic 38 TRANSONIC KIND fUNNEL INSTALLATION Power for the model CRP Is provided using a British erospace ouned Tech Developuent TDIT00 alr motor tnt uch ofthe rig. devclopeent. programme was Concerned with the capability vo. provide precisely controlled power to the individual CRP stages Tests in both an out-of-tunnel tost house at M~ 0 and in the ARA 2.7 mx 2,4 m transonte wind tunnel (to M'= 0.8) have demonstrated that the individual rotors ean be controlled to around £0.1% of rotor speed in spite of the. rotor-to-retor interact ions Following normal ARA practice the rotors of the CRP small seale Installations, on rig or aircraft nodel, “are cleared using’ blade mounted strain gauges. “The outputs of the strain gauges are used Ses means of aasessing rotor dynamic disturbances critical points and flutcer, In addition, the rig 1s equipped with accelerometers and these are used 94 to supplement blade strain gauges and shaft balance data to give overall blade, rotor and rig dynamic data Ik is noted here that It {e necessary, as illustrated in Figure 2, to determine the thrust of the model exhaust nozzle. The use of an underfloor Compressed air feeding Force balance, wherein the air travels into the metric model system perpendicular to the thrust axis, permits a staple treatment of nozzle thrust based on nozzle coefficients. A set of experinents In the Mach Simulation Tank (MST) is used to determine the nozzle thrust and discharge coefficients in a Quiescent exhaust environment. Use of common critical venturi meters (for turbine drive air flow hneasurement) in the alrfeed to the MST and TWT allows the Linking of nozzle airflow accounting procedures to enhance security of the derived ‘ate? 2.2 Development of shaft Mounted Thrust /Torgue Balances Following on from the successful development of shaft mounted thrust/torque measurenent strain gauged balances for SRP application, the Fequirenents for a pair of CRP balances have been Addressed, Key features of these balances were as Follows a) Transmission of balance electrical signals by slipring for the front roter and telenetry for the rear rotor: b) Mounting of balances to shaft and rotor hubs designed to minimise stress ‘signatures’ at the balance flexures, and ©) Thernal problens arising due to heat flow along the CRP ‘shafts to the cold alr motor turbines to be minimised by design, ‘The compact nature of the model led to the overall hub design schene outlined in Figure 4. The attachnents to both the shafts and to the rotors Were specifically designed to produce @ symmetric Tov-stress torque and thrust peth to the balance flexure. sta | ae | ee fl | FRR } FIG 4 RB FRONT AND REAR HUB ASSEMBLIESThe shaft comprises strain gauged balance design, Figure 5, a set of suitably disposed flexures In & basically symmetrical layout with strain gauges Placed so as to provide symsetry and duplication of the key elements, A high standard of accuracy vas achieved using @ layout of multiple conpensat ing gauge placements which were devised to ainimtae the Problems noted In b) and c) above. In particular, the design” ain of a balance giving 2h repeatability in a static calibration was successfully achieved, — Frou Feowees ear FLORES I rowg cats remus causes coun oat Fok He) Hi gl tae OETA igo a owner ern ernie : FIG SARA PROPELLER RALANCE Development of the required telenetry installation to support “the above accuracy target was not without problems. The high CF leads required a specially configured housing for nine telemetry units, Considerable care was necessary to ensure that the telemetry unit installation net exact ing electrical and structural requireuents, Special Composite retention rings were designed into. the telemetry unit. housing scheme and a high (248) Standard of signal processing resulted from the detailed electrical “and structural refinements. jonotry units were used for transmigsion of strain gauged balance signals as. well” as 8 from blade nounted strain gauges. 3.3_Propelte Hubs The high stress levels arising in snall scale propellers necessitate the closest’ attention to detail in ell aspects of blade and hub design and manufacture. Blades are fabricated from selected bar stock titanium toa high tolerance on profile and twist distribution. Key features’ In. the production of model blades are the analysis of the Blade-to-shank structural region and blade sect ions 95 using finite element stress analysis methods to arrive ata geometry giving the required Aerodynanic profiles yet moderate peak stress Tevela.. The studies have invartably benefited from the wide experience of rotor ‘club" partners this” process, especially in respect of the acro/atructural design and im relation to blade dynamic stress assessments. In order to achieve the required blade surface finish to miniaise risks of fatigue failure, blades were hand Finished and Polished after NC machining. ‘The method of attachtent of rotor blades to hubs as an inportant feature of the hub design. In order that any selected eet of blade angles could be created in the front and rear blade rows, a clamping scheme was used to hold and secure the blade shanks." The design of this feature necessarily included shank blend. region to Rinimize "stress concentrations atthe blade/hub/shank Interface region as well as blade retention and setting features, In practice, It hhas boon shown that consistent blade settings to the order of 20.1” have been ach(eved. The availability of an infinitely variable blade angle sett ing/clanping schene has enabled sett ings to be adjusted to correct for errors in predicting the model rotor 'live' twist which arises from the combined aerodynante and dynanic loading of the elastic blades. ‘The design of the rotating components had to allo for dynamic loads, Significant cyclic stresses arise due to operation with combinations of high power and angle. of attack. A relatively Conservative level of cyclic stress has been allowed within the working stress envelope; maximum Allowable cyclic. stress levels were based on factoring of the Infinite fatigue life value In overall terns the design and Integration of blades, hubs, shaft balances, sliprings and Selometry units has represented @ significant ‘engineering challenge The control and survet ance compressed air driven air motors Supported CRP hubs and rotors demands on conputer software systens for test support. Figure 6 shows the overall scheme of computers and date system used to provide on-line Feal-tine agsimilation of model and air motor health! and perfornance data. In addition, the more traditional analogue and dynanic data analysis Techniques are required for ensuring acceptable nargins on blade strest and related rig dynanies of teste using with shate balance wakes significant The overall monitoring and control of the test falls into two categories 1) Continuous on-line Fig/rotor/notor contrel/ health’ parameter monitoring: b) Review of on-line performance data to ensure collection of suitable sets of paranetric perforuance data The category a) data is provided in timescales of 0.1 to 0.2 seconds, whereas the delay tine for plotted category b)' data 1s in the region of 10 Feconds‘Seale Iuolated CRP A typleal isolated CRP operating matrix is illustrated in Figure 7. In general, It is unlikely that the ‘measured CRP performance will natch the predictions, Some of the reasons are: 8) Unpredicted blade root and tip region performance, b) Front to rear rotor flowfleld interference, ©) Non-correspondence of predicted and actual Front and rear rotor blade ‘live’ twist, 4) Deviation of measured from predicted rover effictency variations, €) Lack of completeness of the predict ion computational model, £) Scale effects due to lower test Reynolds nosberss. In the course of producing the CRP performance data fe has been found that the blade setting angles required to. produce a given level of thrust fall Very close to the predictions and that the general form of thrust, power and efficlency vartattons are largely predieeab ‘The actual levels of efficiency are, howaver, found to deviate slightly from the design point predictions; for condit tons far avay From the design point deviations are more marked. ar oeTETs ssouarto | (ae was ae hoa oe c PRIME 236 “Ok UNE DYNAMICS DATA aunt. I ae CONTROL Gueuen Bue ts et Stsie wit Faansouce PresSunes Siecle cae Prouaae ERtEGE Tons 1 i { | 1 1 1 I HP AI ANG ROTOR PaRanERS: | ow, WALtes, I Broo tacticents ror exe |) rotor conrpouen 1 i ood TEST ents | 1 1 1 ! 1 1 ecb focue at reRs ron owe ENrneesng ReveN mo “N-ENE ATA STANDARD eon ee ngeane || wr conrmor adaw "| 1 eal PAINE 9955 ‘OFF-LINE reine Dara TaguLaTion Hat OATA BASE Bore Sete FIG 6 DATA SYSTEM FOR SMALL SCALE CRP TESTS ow otk sol | rhe Bie am _ ‘OpeATNG_TEST_eNvELOMe OVERALL PROPELLER PERFORMANCE RESUATS FIG 7 OPERATING MAP FOR SMALL SCALE ISOLATED CRP 4 LARGE SCALE CRP EXPERIENCE 4 ves of . The most recent and most substantial of the club activities on open rotors har been the RI2 research programme on contra-rotating propellers of 0.8 m Glanecer, The main objectives of this programe 1) Acquisition of CRP performance data for comparison with CFD hased predictions; b) Measurenent of near field and far fteld note ©) Evaluation of a laser based method to neasure "Liver blade twist and camber! 4) Slipstreas peasurenents using rakes placed dounstream of the CRP: ©) Application of laser holographic flow visualisation unnet to CRP in a high speed wind f) Study of CRP flow fields using laser transit anemometry: ) Neasurenont of blade surface pressures using flush mounted pressure transducers (steady and dynamic signals): h) Use of a telemetry system for transmission of all data from roter sounted sensors, including shaft mounted balances: The progranne covered two designs of CRP and there were two separate test series In the transonic tunnel. In the first. of these nolse and performance data vere obtained and in the second Flow field surveys were made. In the following sections the main characteristics of the rig and the results of the programas are outlined.4.2 Motors and Rig At an early stage in the overall propeller research Programme,” the need for a compact, high powered Propeller drive for the transonte wind tunnel was Fecognized, The leading requirements were: 8) Total power in the region 1000 ¢o 1200 sHP (150 vo 900 kD; b) Provision for SRP or CRP capability with power equally divided botween shafts: ©) Long power-on testing endurance for naxinising tunnel productivity; 4) Accurate and rapid rotat tonal speed control (order 20.1%) for SRP or CRP; ©) Long nean-time-between-overhauls and high operational rel lability; 1) Low noise to minimise adverse effects on Propeller noise research; 8) Minimal contamination of wind tunnel ctreuit airflow to retain aaxiqus tunnel product ivity. These requirements wore met by a pair of AC induction motors made by Able Corporation of the USA, fed froma pair of Inverters giving a variable Frequency supply to 360 Hz quaci-AC. The main characteristics of the motors are a) Power pes shaft 660 SHP (490 kW) at 7000 rpm; b) Envelope size per motor 25" tong (0.63 m) by 1a" atanster (0.36 =); ©) Weight per sotor 850 Ibs (385 ke) 4) Integral lubrication and cooling, schene. ‘The motors were first used in the Programmes on single rotation propellers}. ERP programe, “they were incorporated specially constructed Rolls-Royce tig (Rig 140) Re and RO For the This rig {s shown Installed in the transonte tunne! In Fig. 8. The upper diagrans show the rig Installed together with the acoustic liner for the conbined performance and nolse measurenents of the First test serfes; the lover iagrans show the installation In the baste tunnel for the combined performance and flow field survey of the second AS'shown In the lover diagram, the two Able motors tere wounted In the rig in tandem, A structural Sleove as used to carry the motors ahead of @ Fobust sting, the sting, {9 Curm, was pivoted from Elarge strut which traversed vertically so. that, fs propeller angle of attack was varied, the CRP Femained in the centre of the tunel working Section. The CRP nacelle and rotors were developed by Rolls-Royce with support in respect of rotor design and rig integrity from Dowy Aerospace and ARA,. "The components of the CRP asseably consist of a) the drive shaft coupling nodule; b) the CRP module front and rear rotor elements; ©) the non-rotating nose bul let 4) an extensive rig health and performance instrunentat ion system. A conaxtal system of three shafts connects the Fotors and bullet to the Able electric motors; the inner stationary shaft 1s used to prevent rotation of the bullet although bullet structural support Is, piven by the CRP module. - pe: RP IN AcoUST NER Nase teen Sa ote ue gee ines (ee fio swat ae —] FIC 8 CRP MEASUREMENTS DURING TESTS IN 97 = Se TASER TRansir aewoneraY ACOUSTIC-LINED & BASIC HIGH SPEED TUNEL4.3 Rig and Model Instrupentation ‘The primary Instrumentation of the rig vas a palr of shaft mounted thrust end torque balances and a Substantial set of surface static pressures to enable rotor performance to be evaluated, coupled wich ‘extensive rig and roter “health instrumentation including, in particular, strain gauges on the rotor blades Im addition, a set of blade mounted (flush) Pressure transducers was provided for assessnent of dynamic and steady state pressure variations (Figure 9); also, a system for seasurenont of front and rear rotor blade ‘live’ twist and camber was evaluated. These measurenents will contribute to a fore thorough understanding of the coupartsons between predicted and measured CRP performance The ‘live’ twist measurements may also be used £0 improve the aero/structural prediction wodels which are necessary for realistic. prediction of aerodynamtes ‘The tost programme included an assessment of the Likely effects of the exerescences on the rotor blade surfaces due to the placenent. of pressure transducers and strain gauges (and all associated wiring): in the first test series a set of totally clean blades was included for performance and noise measurement purposes The acoustic measurements of the first test series were made on a set of fixed microphones and on a hobile plcrophone array mounted on a slerophone traverse rig as shown In the upper part of Fig & For the flow field measurenents of the second test series, a set of three downstrean rakes. provided surveys of slipstream total pressure, swirl angle and total temperature and, az shown In the lower part of Fig 8, laser transit anenonetry and laser holography eystens (see section 4.4) wore deployed The overall data system was similar to that used for spall scale CRP work, as shown in Figure 6 honever, a number of additions were embodied for the large scale work. Figure 10 shows network of Signals and monitoring Funct tens. the R12 ‘The leading Facets of rig/nodel health and contro! were as follows analysis, 5) Dynaste of blade strain gauge pressure transducer and accelerometer stgnale, on-line and using recorded sequences: b) Monitoring of test management computer data covering bearing temperatures, Iubrication system and cooling system performance, on-line: €) Assessment of real-time Able aotor/Cort ina inverter data to. ensure compliance with, established operational requirenents; 4) Survettiance of convent fonal forces and pressures data signals using established rnethods 98 7 PRESSURE TRANDUCERS (ON SUCTION SURFACE FRONT VIEWS OF FRONT. ROTORS > STRAIN GAUGES ON PRESSURE SURFACE | FIC 9 SCHEMATIC SHOWING LOCATIONS OF PRESSURE TRANSDUCERS AND STRAIN GAUGES (OW BLADE TYPES A & B €) Provision of colour graphics displays of a menu-controlled. set “of selectable multf= parancter plots: 1) Control of computer driven _euto-traverse equipnent for microphone noise Tield surveys; f) Network of audio Hinks to provide conmuntestion between the test manager and funct fon specialists together with a multi-lecation set of conputer monitors displaying key test ‘As Is ovident from Fig 10, the task of overall test control was a complex one. At the planning stage, detailed attention was needed to all aspects of the task, particularly to Interact ions between systems ‘and to coordination betwoon system operators. A large team of specialists rom Rolls-Royce, Dowty Aerospace and ARA carried out the teste and the success of the progranme oues a great deal to their highly disciplined teamworkNOSE MEASUREMENTS CONSOLE ONLI. STU OF no‘se ata Faw TBA AUTO TRAVERSE BT as} V4G00 TO-vemey coLbection on PeMOPROATE TRAVERSE OTA tenacrve procaan sorTie| Ponmaaie Sae Gre tm Horan Bie (apmiarin Coons systes TALE ELECTRIC oroRS: FRONT/REAR vena egeinay SeSanare caraste Suart I ANALOGUE OATH ACOUSTTION SYSTEM BECOMES OK RAGHETIC TAPE ALL ynartc OUTPUTS BIRBE SrRaN caved’ PRESSURE TRANSDUCERS IeceCeRO [TeLeMeray ConnrTiONNG wWiTS WAVE FR ANALYSERS vente Sisto, innate IR WARAGER CONSE RTeRACTVE pur FoR AVERSE ReGURENENTS ELECTR MOTOR DANE CORSTLE “argos ron | Moron ano foro cole | Weve eUugweamion | FAUCET ex curr BREAK | coun Speco a wr | Sean fonreats’” | Tees oToR HEART MMTonNe | none reves BEARING TEMPS, oro vous oR RENT { ‘ABA DATA ACOUSITION SYSTER Taxosrone ene | 200 comune ich spe 1 ieanotoures, nibirace | SALES Fig HEALD CONDITION Beatin | SARS em | Beas | a] Ge” | es FERRERS TEST NANAGEFENT COMPUTER RECORDS AND OIPLAYS open HEALTH PARAMETERS, BExao Tens le system AN COLLIN SiSTem ORI COLOUR GRAPIC DISPLAYS SHORING PropeLLeePenronMAnce ORT Checpsta. ACH, RPM, ETE HLT SeLEcTagie ScaeENS Fo | SURVECLANCE Oe ALC arabe TERS BRA aR TEL Aihletns cise TAL KEY TEST Panam ens win ew Sececrion FIG 10 R12 PROPELLER RIC CONTROL, MONITORING & DATA ACQUISITION SYSTEM44 sud a were used to large seale RP Three separate laser techniques provide research data during the test programme; all three techniques were created and ‘directed by Rolls-Royce research team enbers u Twist The aerodynanlc and centrifugal loads on the CRP blades cause the bladss to deflect and twist, altering the pitch angle, aerodynamic loading and Possibly sectional shape. To astess the magnitude Of the twist and shape changes and thereby to assist with assessnent and. interpretation of Beasured to predicted performance variations, a Taser system using an optical trlangulatton approach was used ‘The system measured blade shape for each of the blade rows (front. and rear) at three selected radial stations. Previous experience of Rolls Royce on turbofans had indicated that blade shape could be measured with an accuracy of a few thousandths of an inch. Interrogation of the data suggests that high accuracy was also obtained under the conditions imposed by the CRP transonic tunnel test requirenents. The Indications are that small Variations in blade twist could be measured and that changes In blade camber could be observed. Lager Ancmoneery A key part of the Phase 2 tests in the basic THT involved the use of a long throw two-spot laser translt anenonetry (LTA) optical system. The LTA technique has the advantages that local velocity (Speed and direction) can be measured by a non-intrusive approach and that quantitative information on turbulence TevelS can be obtained Disadvantages are thet the method requires significant elapsed tine to collect the data and that an ofl-aist flow-seeding aystem Je required to provide particle paths O11 mist particles passing through the zone Into which the laser ‘beans are focussed create Scattering of the light bean; these variations of Tighe level are recorded using photomultiplier and a correlation of the derived signals peraits evaluation of the velocity. The method uses tvo pairs of focussed beans, the orientation and position of the mecting points of which can be Fotated and translated to enable velocities to be determined over a volune of the flow field. A spectal laser cart, mounted In the tunnel floor, was used to carry the LIA equipsent including the Argon ton laser and a largo mirror and related traversing equipuent. The data processing and control of the laser vas established inthe TT control room. Views of the laser cart and typical beam paths are shown In Figure 11. To relate. the Taser’ instantaneous measurements to rotor blade position, a precisely tined signal was created Using a transverse pair of HeNe laser beans, The Flow vas seeded by plunes of 0.5 jm particles of nineral ofl injected from a special strut erected In the TWD settling chamber 100 ‘The LTA syston was used to collect data for vartous locations in the CRP flow Field including: a) Overt ip vortex region b) Inrotor (between blades) region ©) Retwoen rotors region, 4) Upstream and downstream of the rotors. ‘The capability to generate substantial sets of CRP velocity field data using the LTA system was Genonstrated for both high and low speed. test condi tons Jor Holographic CRP Fi. ‘A non-intrusive double-pulse laser holographic flow Visualisation” system vas used to produce Interferometric holograns of the flow in the large Scale CRP tests®-7.~ The bean from a high power pulsed laser (class IV) was used to illuminate the CRP podel as shown in Figure 12. The method offered the advantage of being relatively rapid and It was practical to create 80 holograns ins single run ef ‘the order of one hour duratton, Including GRP power level changes and laser resetting. _ | — ilbh ee . a My Fie 1 LASER CART WITH OPTICAL HEAD FOR LTASafety was keenly observed during all taser operations and an Incident free series was achieved. Alignments of the optical system wore perforned using low pouer lasers. ‘The LTA and laser holographic systens were designed s0-as not to create mutual Interference A large musber of valuable holographic images of the GRP flow Tleld were produced and a number of significant features have been observed Including. 5) Tip vortex path and chopping of front vortex by rear roter: b) shear layer in blade wake region: ©) Supercritical Flow features including expelled blade leading edge bow shock. Interpretation of holograms was aided and enhanced by a closed cireult video system ape a wove. fas Pease Cress agaer TWT WORKING SECTION View FR UPSTREA FIG 12 ARRANGEMENT OF LASER HOLOGRAPHIC SYSTEM IN THT WORKING SECTION 4.5 Development of the Acoustic Liner Genoral Reauirenents ‘An acoustic evaluat lon of the basic ARA tunnel vas conducted in 1985 using 2 munerical model; this Study Indicated that the tunnel in tes ‘baste’ form Was unsuitable for acoustic measurements due to the Strength of the reflections from the hard 22 ft (6.7 m) diameter pressure shell. This finding was confirmed by 2 brief set of measurements inthe wind tunnel and. was consistent with prior US cxperience® which had shown that lack of acoustic treatment could give rise to errors of the order of 5 dB in peak noise level Following this evaluation, programme was launched to develop a acoustic Hiner for the tunnel, The development, Feported more fully in Refs 9 & 10, required both aerodynamic and acoustic problems te be solved UK ‘club! supported Fonovable Aerodynanically, the acoustic liner was required to euhIbIe the following characteristics 101 section cross ‘towable a) Maximum fect ional model scale. practical working inensions to maximise b) Working M range from 0.2 co 0.85 with the main emphasis in the region 0.6 to 0.8 ©) Uniformity of Local Nach number to £0,005 in the rogion of the propeller Acoust ie requirenents were: 8) Attenuation of incident sound waves over a wide bandwidth, expecially in the propeller plane, to enable’ free field sound measurements to ai accuracy of #1 dB oF better; b) The absisry axial, redial and clreunferont ial to make noise measurenents along Tines. ‘The second requirement led to the development of a cosputer controlled microphone traversing rig System in parallel with the development of the acoustic Tier: Acoustic Design of the Liner A munerical model was created by Rolls-Royce of a Gound field in the wind tunel. This model was ised to represent various wall boundary condit fons in corbination with sound fields of prescribed directivity. The numerical sede! was based on decomposition of the sound flelé Into plane wave rays; the direct and reflected field strengths were sumed at x selected set of ‘observer’ locat ons Te was further refined using the concept that multiple reflections of the source sound field ould arise and that the strength of the observed Teflections would depend upon the positions of the Source and observer within the tunnel and the reflection coefficients of the tunnel walls. Subsequently, the model was extended to include the influence. of freestrean. Mech” nomber and the refraction of oblique incident waves by the tunnel wall boundary layers. Varlous source direct ivity patterns, were considered Including simplified hodels representative of SRP and CRP open rotors. Using this mumerteal model, design studies were carried out to establish an acoustic treatment hich would be an acceptable conpromise between achieving good sound. absorption Gall thickness of the liner to a practicable ininum. The requirement that the depth of an Scoustic bulk absorber” should be at least one Guarter of a wave length for the louest frequency Of concern lod to the selection of an absorber depth of 6 inches (0.15 m), corresponding to a wavelength of 2 foot (0.60 m) which vould be obtained at » frequency of 570 Hz and a local speed of sound of 1140 ft/sec (347 m/s). To obtain a Wide bandaldth of scoustic attenuation, an open cell polyurethane foam was chosen as the material for the bulk absorber ‘and keeping the The Inner airswept surface of the Hiner was made of perforated. steel sheet with an open area ratio Close to 37K, which was predicted to be essentially Cransparent acoustically In the frequency range of interest. The chosen depth of absorber reduced the working section of the tunnel from 9" x 8" (2.7 a x 2.4 my to 8! xT! (2.4m x 2.1 m) approximately.cts of Liner Des! In aerodynamic terms, the design of the Liner represented a significant challenge, given the Aforenent loned stream uniformity requirements Joint computational /experinental approach was used to generate a suitable belinouth Inlet contraction design, to define longitudinal area variation, and to develop a diffuser arrangement compatible with the presence of the bulky model support strut Ahead of the R12 progranne, because of concerns regarding the high blockage of the rig in the Tiner, a set of TXT tests war made using @ dummy rig in the tumel wleh lner; these tests also included an Snvestigation of the Influence of the ovement of the sicrophone traversing rig on Tongitudinal velocity distribution The computational sade! (SPARV) used to design the Hiner vas also used co support the analysis of measurements taken during these testing phases bal 2130 ierosen BELLMOUTH IME ae CONTRACTION POROUS BACKING poate BaRa sree, ‘Quaonan oF 20m, couse LNeR ACO Les rare TRARATWT ~~ ranysvense | ("Riess sete Ly sasic rat was site oi, bs are SSaLuMINi uy a i in 4h 23109 Sat ge rons FIG 13. DETAILS OF ACOUSTIC LINER FOR ARA TRANSONIC TUNNEL ‘The mechanical design of the Liner and its location with respect. to the wall” of the, tunnel plenum chanber is Illustrated in Fig 13. The ner. has been bu(lt ax a free-standing rectangular’ shell Torned so as to retain the essential acoustic properties at the walls and co include an inlet bellmouth contraction and an outlet diffuser section. It fits closely within the existing TWT porous (22K) walls and provides a 12 ft (3.7 m) length of acoustically treated surface, which tz considered ‘sufficient to cover all Ilkely model test requirements. A boundary. layer bleed. is provided to suppress separation at the corner 102 between the Inlet contractfon and the TWT walls. The foam is carried between longitudinal rails which serve as reructural members and to which the Inner walls (37% porosity) are fixed using discrete doned-head fixations to local hard points in order to mininise sound roflections. The rear wall of the liner is ventilated, for structural and aerodynamic rather than acoustic reasons ‘The wind tunnel eiroult has a removable section, innediately downstream of the test sect fon, which is drawn back to enable model support carts to be interchanged, The sane process provides access for inserting. the acoustic liner “into the tunnel, installation requiring slightly more, and renovai rather less, than a working day Acoustic Liner Proving Trials Aerodynamic calibration of the Viner in the THT was carried out in three phases, In the first phase tests with a long static probe on the axis of the empty liner established the flow uniformity in the Miner and. were used to develop procedures for tunnel “setting and operation and for calculating Mach number in the liner. In the. second, the effects of the microphone traverse rig snd a seall Propeller rig were established and measurements of Propeller performance were shown to match results fn the standard (unl ined) tunnel, The third phase, which Involved tests with the dunay R12 rig, led to Aerodynamic modifications to the liner diffuser hich restored Jost power margins and enabled the Tunnel to be controlled up te a Mach nunber of 0.8 with the RI2 model Installed in the liner. The calibration showed che freestream Mach number in the acoustic liner to be predominantly a funetfon of the standard tunnel power setting parameters, fan speed ote, and secondarily Gependent upon the level of plonun chamber suction; furthermore, the presence of the thrust producing CRP was expected to be significant. The Tiner Mach urber' calculation routine was based on the conventional approach to model blockage and included an allovance for tunmel wall constraint as well as model support Interference and the effect of the presence of the microphone traverse rig, Tests have systonatically included measurements of wall static pressures as well as upstream and plenum chanber pressures The acoustic development of the Hiner9s19, carried out under the #7 ‘club’ progranne, involved a Series of tests with propellers In the lined TWT, backed up by acoustic trials with the Tiner outside the tunnel, Propeller tests in a low speed anechole tunnel at’ RAE Farnborough provided a Valuable datun The wicrophone traverse gear enabled longitudinal traverses to be made at a given radial distance Tron the axes but at different angles of azieuth (and therefore with different Intensities of sound reflection from the walls). Traverses could also be made at the same azimoth but different radial Gistances The flret trials of the acoustic Liner showed unacceptable reflections from sone features within the liner, The reflections were manifest as Gitferences in feld shape at different azimuth angles and as. departures from the inverse square Taw along a ray path. By an iterative process,which Involved significant changes to the nechanical design of the liner, the sources of Feflection were identified and eliminated. The Final trials in the R7 progranne, which were with 2 Single rotation propeller of 15" dianeter, showed that a high level of anechotetty had been achieved Ae Mach “nusbers less than approximately 0.6, Inverse square law behaviour waz observed In the far field and directivity patterns measured at different azimuths watched ‘each other closely, Above M= 0.6, reflections from the wall became increasingly obtrusive, particularly in the forvard are. Despite these reflections, attributable to boundary layer refraction, the Tiner vas considered to provide adequate measurenents of the key noise paranoters at’ Mach” numbers characteristic of Propeller aircraft at cruise 4 format 2 and As already stated, the tests of the large scale CRP {in the ARA Eransonie wind tunnel were undertaken in two phases, vizt Phase 1 (MAY/JINE '89) CRP performance and noise measurements using Microphone Traversing Rig (MTR) and shaft balances in the TWT with acoustic Hiner fitted Phase 2 (OCTOBER 89) CRP flonfield surveys with performance measurenents using Taser Frangit anenonetry and laser holographic techniques Prior to the First tunnel tests, however, an exhaustive check-out of the rig, ts anctilary systens and Its instrumentation was carried out in the test cell of the ARA Propeller Test House (PTH). The PTW was built as an adjunct to the overall propeller programme, to” provide the capability, outside the tunnel, of exercising Propeller rigs and thelr aystems over as wide an operating range as possible at nominally static conditions, It has proved Invaluable in the final, pre-tunne! stage of development and proving In ali the rotor progeasnes Rotor Blade Stress, Vibration and Flutter Clearance Although the proving trials in the PIH provided information on rig and rotor dynamics over the range of conditions achievable at nominally zero Forward speed, the full dynanic clearance of the RI2 rig had to await the first entry into the Eransonte tunnel, Thus, at the outset of the Phase 1 tests In the acoustleally Iined tunnel, It wae necessary to "elear' each of the rotors from the Viewpoint of stress, vibration and flutter. Whilst critical “rotations! speeds are comparatively predictable from established rotor natural Frequency (flap, torsion) data, the cecurrence of classical flutter ts far less predictable. The prinary variable in classical fs helleal tip tach munber. flutter assessments Tests to clear the rotors were performed by a carefully monitored sequence of rotor and tunnel Speed variation az illustrated In, Fig 14, Ata Blven tunnel Nach number, the speed of each rotor in turn was progressively Increased uti! the Talt of the test envelope waz reached: rotor speed was then reduced, tunnel Mach number increased without exceeding the safe helical tip. Mach nusber established at the previous tunnel Mach nuaber, and the process repeated. The flutter clearance costs were carried out for a single setting of the rotor blades and so provided a good first assessment Father than full clearance of the rotor. Blade Stress and. vibration levels were therefore nonltored carefully through the entire programe ys HA AC ys HAL TP MAC T < Vea te oF Gk Oh OT oe ay Pe Poaceae, “unt ud va ON-LINE DISPLAY FoR STRESS AND VIBRATION SURVEY AT FIXED ROTOR BLADE ANGLES Fig 14 Performance Testing The specific objectives of the performance test ing were as follows 4) Determine efficiency at key take-off, climb and cruise operat ing points: by Exanine a range of torque splits; ©) Examine speed differentials at constant thrust and power 4) Determine efficiency and general performance over the Mach number range of 0.2 to 0.8 “TRUST COEEFIOENT ca POWER COEFICENT FIG 15 _ OPERATING MAP FOR LARGE SCALE CRP AT CRUISE DESIGN BLADE ANGLES & MACH NUMBER 103The case of spood control of the Able electric notors made i¢ practical, in short time, to examine a satrix of front and rear poser settings. Figure 15" shows the mecsured variations of thrust and power coefficient for the swept blades set at the established cruise blade angles. It ts apparent that the performance of the CRP can be assessed over a fine mesh of the front and rear rotor advance ratios. From tests covering a sweep of vartables, as illustrated in Figure 15, specific levels of power and/or thrust coefficient were selected for further Study. using the nolse survey equipment. This approach was very economical of tunnel tine typically, the performance points were collected at between one and two points per minute, while nolse traverses usually took an order of magnitude longer As Fig 15 shows, the thrust and poner coefficients exhibit typical CRP characteristics, nasely that the front reter directly influences rear” rotor coefficients uhilst the rear rotor has relatively little influence on front rotor coeffictente. By basing the pattern of testing on this recognised charactertatic, risk of exceeding rotor blade Stress limits were reduced. ‘The measured CRP overall efficiency 9 (for conbined Front and rear rotors), derived from tests across a nutrix of blade angle ‘settings and front and rear advance ratios, for take-off free stream conditions, 1s plotted In Figure 16; the plot shows 7 variations with overall ‘thrust loading. The Envelope of data shows the optimun value of overall efficiency that can be achieved for a given overall thrust leading. The ‘opt {mm overall efficiency! curve provides a useful basis for making comparisons for different roter speeds, "blade angles and rotor designe. ova, Seri 1 OPM OVERALL EFFICENGY 4 0% - . Ts + , 2 bx "HUST LOADING FIG 16 VARIATION OF OVERALL EFFICIENCY W1THL THRUST LOADING. SWEPT BLADES AT DESIGN TAKE-OFF ANGLES 104 sr Lonel Ra ovP0! primum : ; , veRAKL EFFICIENCY] Ww ‘SWEPT BLADES M=OTS SALES OFFSET 2% FOR CLARITY aT 0% STRAIGHT] BLADES. ee SLATE ANGLE oT PRR IN DESK IN DESIGN 125, IGN-25 DESION 1.25 oy oa ‘THRUST’ LOADING ‘THRUST LOADING PERFORMANCE OF STRAIGHT AND SWEPT BLADES [AT HIGH SPEED CRUISE CONDITIONS Fig 17 Comparative ‘opt imu overail effictency’ data for wo of the rotor sets ts. shown in Figure 17 for a ange of blade angle settings (Bp', 6g). tls Clear that, at the higher cruise Mach numbers (0.75 to.0.8) the changes in Blade angles have relatively nor effects on the optiaum overall efficiency, Ghereas a significant variatton with thrust. loading is generally apparent. Ax expected, the swept Blades achieved @ higher efficiency at the crulse Mach levels. than the straight. bladed CRP” when operated at realistieaily high thrust. loadings ‘An examination has been made of the reliability of the shaft balance data; this study was conducted in Conjunction with the Phase 2 laser anenometry work The extended set of data samples at a nominally Fixed set of test conditions (Mach number, advance ratios and blade angles) are shown in Figure 18 for thrust coefficlent, power coefficient and cffictency; the data was collected during four Separate tunnel runs of durations ranging from 20 to. 150 minutes. The overall spread of the front and rear rotor data samples (s of the order of one per cont for efficiency and 1 to 1? per cent for thrust and power coefficients. On this basis, back-to-hack testing should provide satfsfactory discriminat ion of performance increments.REAR ROTOR] EFFICIENCY EFFICIENCY |?" Pl | g ‘ 0 TSO rnin 2 100 OTe nn) 202 T THRUST THRUST oy |} — COEFFICIENT —]———- mbt fetal a ke Q 0 0 Src mnt =O Fa wo 15 wa? — POWER POWER COEFFICENT [ eee ° 0 100 Worm 20 0 30 0 "S0rme tain 200 FIG 18 EXTENDED DATA SAMPLES AT FIXED OPERATING CONDITIONS SHOWING STABILITY OF SHAFT BALANCE MEASURED PERFORMANCE. PARAMETERS 105Noise Measurements 30” SIDELINE In the course of the Phase 1 test entry, noise ERE ne) survey traverses were conducted for straight, swept (instrumented) and clean swept blade designs. The Fangs of conditions covered freestream Mach nunbers from 0.2 to 0.8 in combination with @ wide range of roter blade angle settings (2p, Gq) and various power settings: Fig 19 shows the microphone traverse rig and the positions of the traversing microphones, and also the fixed wall and blade sounted microphones, im Felation to the rotor. In. the configuration illustrated, ‘the rig carried two pairs of lcrophones’ located at 19 inch and 30. inch ‘Sidelines (0.63 and 1,0 rotor diameters) which were traversed axially to. provide noise field directivity, power and Frequency spectra, A second configuration, with a single microphone which could be positioned’ anywhere within a cirele of $4 inch Gtaneter centred on the rig axis, was used for detailed measurenent of Fadial and azimuthal variations of the sound Tield. In practice, it was shown that close pass traverses could be performed fo within 0.08 rotor diameters of the biads tips | NI Be |r oe " ia 16 19 MICROPONE LOCATIONS & wR TRAVERSE PAIS As noted inthe previous section, noise easurements were mide at selected points on the operating matrix of Fig 15. Sone typieal data are presented in Fig 20. Each graph gives the results For a pair of microphones at the same sideline Aistance but different azimuths, showing the field shapes given by the Individual ‘microphones and by thelr power-averaged sean. AL a tunnel Hach number 106 ran allen 19 SOELNE Carrier! . eee! oun anda! FIG 20 TYPICAL CRP NOISE DATA FOR Low & HIGH TUNNEL SPEEDS & SELECTED FREQUENCIES of 0.2 and at blade passing frequency, the results Show very close ‘agreement between the t90, icrophones at the 30 inch sideline, indicative of the high degree of anechoicity achieved by the Vner, Ata tunnel Mach nuaber of 0.75 at blade passing Frequency, the microphones at the 19 inch Sideline show close agreenent in the region of peak nofse but evidence of reflections in the forward and rearward ares. At 3. times Blade passing Frequency these reflections appear slightly stronger but the agreement between the” t¥o icrophones at the noise peak 1s still reasonably god. The more detatled radial and aziauthal traverses confirmed this general picture and reinforced the conclusions reached!0 at the end of the small scale SRP trials in the acoustic liner. In particular, it wns found over mich of the Mach number range that the azimuthal variation in the sound field was small and that there was a significant region of the field in which the Inverse-square decay law was satisfied6_coneupne REMARKS. A significant set of advances in wind tunnel technique for the study of high speed contra~ rotating propellers has boon achieved at ARA Bedford, 6.2 The techniques for small scale isolated and installed CRP rotor systems which have been developed wil] make «worthwhile contribut fon to the overall study of propeller Installation technology. 6.3 Methods for the study of CRP designs at large model scale have been developed including performance measurenent, noise Field measurement, and flow field measurement Visualisation techniques. ‘and Acoustic measurements with the large scale CRP have confirmed the earlier conclusion, from tests with a shall scale SRP, that the acoust le Hiner Is effectively anechote re Jot The oxtensive and invaluable support of all UK open rotor club menbers, BAe, RR and DA, as well as the DIT and our colleagues at AFA Led, 1s gratefully acknowledged herewith, Specific mont ion must be made of the Liaison of RR, DA and ARA In respect of the large scale studies reported herein which were also supported by RAE Specialists and MOD contract AE/124/89, ‘The small scale studies Included contributions fram BAe and DA specialists as well as the use of tho Bas owned 101700 air motor. 1 Beavis DG, Poznlak OM, "Factlities for the development of propellers and propeller installations at ARA', RAeS Conference Advanced Propellers ahd their Installation on Airerafe, Paper 8, Sept 1988 107 Harris AE, Paliwal KC, ‘Civil turbofan propulsion system Integratfon studies using powered testing techniques at ARA Bedford", IAA 84-0593, 1988 ESDU, ‘Thrust and drag account ing, for propeller/alrframe Interact ion’, Item No 85017, Nov’ 1985; Davingon I et al, ‘Optical fan blade profile easurenents In fotating turbofans", ASME 90 Turbo Expo, to be published. Haxell CF, ‘Laser anemometry in rotating transonic flows", 6th International Conference fon Photon Correlation and Other Optical Jechniques in Fluld Mechantes, Cambridge 1985. sly Parker RJ, Jones DJ, 'The use of holographic interferonetry for turbo sachinery and evaluation during rotating tests" Journal of Turbomachinery, Vol 110, pp 393-400, 1988. Parker RJ, Jones D J, ‘Holographic flow visual sation In rotating transonic flows! , Institute of Phystes Conference Series No 77, pp 141-146, 1988. Glover BM, Plunkett E 1, Sincox C D, "Noise testing of an advanced propeller in the Boeing transonie wind tunnel with and without test section acoustic treatment’, AIAA Paper no 84-2366. Wood ME, Newsan DA, ‘The design and development of an acoustic test section for the ARA transonic wind tunnel", Royal Aeronautical Society’ paper, Sept 1988. Wood ME, Newnan D/A, ‘The design and commissioning of an acoustic liner for propeller noise testing in the AEA transonic Wind tunnel", SAE Aerotech "89, €398/25, Engine land Instrumentation, 1989