On the Design and Test of a Low Noise Propeller - CAFE Foundation

Report No. 1761Bolt Beranek and Newman Ino.ON THE mSION AND TEST OF A LOW NOISE PROPELLERAn extensive review of noise and performance of generalaviation propellers was performed at MIT" Research was done inthree areas: A review of the acoustic and aerodynamic theory ofgeneral aviation propellers, wind tunnel tests of three onequarterscale models of general aviation propellers, and flight.test of two low noise propellers.One of the two low noise propellers was designed and testedat MIT and has already been described2. The second propeller wasdesigned at MIT and flight tested at OSU. The design and testingof the second propeller is reviewed herein.This rpport deac ribes the general aerodynamic considerat ionsneeded to design 8 new propeller, reviews the design point.analysis of low noise propellers, and finally, compares thepredicted and measured noise levels.To design a fixed pitch propeller for an existing aircraftthere are two points to consider. First, the propeller must bematched to the engine, Second, the power result lng from theengine propeller combination must be matched to the needs of thealrcraf t, Fig. 1 lllustretes this problem,In Fig. 1, the vertical axis is power and the horizontalaxis is aircraft speed. The solid black line represents thepower needed to overcome aircraft drag at any particularvelocity. This power is given by the expression:

Report No. 4764Bolt Beranek and Newman Inc.Table 1Aircraft CharacteristicsParameter Value Comment13.44 M' 144.6 ft2 wing area-05 ,skin frictioncoefficient7. 415 aspect ratio0.65 span efficiency9029.9 N 2030 lbs (utilitywe igh t )The heavy dashed line in Fig. 1 re2resents the poweravailable from the production engine prbopeller combination, withthe engine operated at full throttle. This power is the productof the propeller thrust times the aircraft velocity. The productionengine propeller power was deduzed from two observations.First, the fastest sea level veloci4;y attained by tht aircraftwas 63.2 M/S (141.4 mph). Second, the best rate of climb occursat 40 M/S (90 mph) and is approximately 4 M/S (800 ft/min).Wherever the power available exceeds the power needed to overcomeaircraft drag, the aircraft can climb. The velocity at which thetwo curves intersect is the peak forward velocity.The acoustic design point is the full power flyover at 305 m(1000 ft) altitude which is specified in PAR 364. The acousticgoal Is to reduce the propeller noise, as measured by a groundobserver, with no loss in aerodynamic performance. Hence, thedesign point is to match the propeller performance during a fullpower level flyover. The procedure is to design a propeller

Report No. 1761Bolt Beranek and Newman Inc.which minimizes the induced losses at the design condition.There are six parameters needed to specify the design of aminimum induced loss propeller. These are given in Table 2,along with the associated values for the existing propeller.Table 2Propeller Design SpecificationsParameter Ex1 s t ingPropellerAirDensity 1.225 KG/M~Velocity 63.2 M/SRotation Rate 2700 RPMPower 122 kw (164 hp)Blades 2Radlus .97 M (38")NewPropellc A-Comment1.225 KG/M~ Sea Level63.2 M/S Max. Velocity2700 RPm Max .RotationSpeed122 kw (164 hp) Max Eng. Pwr.3-87 (34")The implied constraint is that the efficiency (85%) beapproximately the same for bath.This is discussed in greaterdetail in the next section.The measured properties of the new and ./existing propellnrs demonstrated that they were identical duringa full power flyover.However, as evident in Fig. 1, it is important to maintainaerodynamic performance of over a range of airspeed.procedure used to analyze this performance is indicated in Figs.2a, 2b, 2c. Fig. 2a represents the measured engine power versusrotation speed on the OSU aircraft.ximate equationP122.3 + .0447 (X-2700)TheThis was given by the appro-

Report No. 4764Bolt Beranek and Newnaan Inc.where P is the engine power in kilowatts, and X is the engineRPM. One point to note is that the measured peak engine power122 kw (164 hp) is less than the rated power 134 kw (180 hp).Fig. 2b and 2 c represent the calculated propellercharacteristics. These are given in non-dimensional form sspower coefficient (Cp = P / van 3 R 5 ) and thrust coefficient (CT =T / 4'"~ 2 R 4) versus advance ratio (V/RR). Here R = Bn, n is therotstion rate and R the propeller radius.The procedure for constructing these curves is given inRef'erence 5. Several general points should be made. First, thepropeller is designed to support a load distribution whichminimizes bhe induced loss for the conditions listed in Table2. This procedure gives load distribution which, In turn,specifies the product of the chorZ times the section lift. Airfoilsections are chosen to operate at the maximum lift to dragratio for the condition. In this instance the sectionc are NACA64 series airfoiis. Given the section lift, the chord and angledistribution versus radius is determined. The inf luence of theaircraft nacelle is accounted for approxin.ately by assuming lowerinflow velocities near the hub.Once the geometry of the propeller is specified, and theairfoil section properties tabulated, it is then possible toconstruct the curves given in Fig. 2b end 2c. One parameter isabsent from Fig. 2b and 2c: the propeller Mach number. The Machnumber is important because the airfoil section properties dependon Mach number. To accomodate the effect in an approximate way,Figs. 2b and 2c are constructed for a propeller rotating at 2700RPM. A more accurate analysis would require a series of curvesto be constructed for each Mach number. This is more detail thanneeded for a preliminary analysis.

Report No. 4764Bolt Beranek and Newman Inc.To synthesize the engine-propeller curve of Fig. 1, followthe steps listed in Fig. 2. First, select rotation rate from theengine curve; this gives an engine power (Fig. 2a). Qiven thepower and rotation rate, find the advanced ratio at which thepropeller power absorption matches the engine power (Fig. 2b) ;this gives the aircraft velocity. Now, from the advance ratio,find the thrust produced by the propeller (Fig. 2c). The productof this thrust times the velocity, plotted against velocity,yields a power curve similar to that plotted in Fig* 1.Design Point AnalysisTable 2 shows that the new propeller differs from the oldpropeller in that it has three blades and a reduced diameter.The strategy is to combine two noise reduction methods in such away that the efficiency is maintained. This strategy is based onmethods developed for reducing noise on the MIT aircraft 6 . Thisaircraft is similar to the X U aircraft in that the propellerdia~eter and rotation speed are identical. The largestdifference is that the MIT aircraft has a 150 hp engine, whereasthe OSU aircraft has a 180 hp engine. However, the similaritiesoutweigh the differences and the same design charts used for theMIT aircraft were used for the OSU aircraft.The goal is to minimize the peak dBA levels recorded by aground observer as the aircraft flies a level path at an altitudeof 304.8 meters. The test condition must be in accord with thosespecif led in FAA Advisory circular 36-l~(a). Appendix F, Section36.111, paragraph 6 states "Over flight must be performed atrated maximum continuous p~wei*.'' In accordance with this regulation,the design point is top speed at full power at the red lineRYM. The maximum sound level on the ground occurs when the

Report No. 4764Bolt Beranek and Newman Inc.aircraft is 5O past the zenith. This is the point at which theBound is computed in the design charts.In Fig. 3a and 3b the effect of changing the number ofblades is examined. Fig. 3a illustrates the effect on propellerefficiency. The figures were constructed by designing a seriesof minimum indiced loss propellers. As the number of bladesincreases, the efficiency increases and both the sound leveldecreases at a slower rate than the unweighted sound levelbecause the blade passing frequency increases with the number ofblades.The noise is changed by altering the source strength.Compare a two and three blade propeller, each of which absorbsthe same power. A blade of the two blade propeller must provide1 v2 times more force than a blade of a three blade propeller andmust have, approximately, a chord 1 ?$ times as large. Becausethe peak sound level is dominated by conditions on a singleblade, this fact means that the source strength for both thethiclcness and loading terms must be larger on the propeller withfewer ~~ades.In Fig. 4a and 4b the effect of aitering the radius isexamined. Fig. 4a illustrates the effects on sound level, Fig.4b illustrates the effects on propeller efficiency. (Once again,the figures are constructed by designing a series of minimuminduced loss propellers. )The horizontal axis is the propeller radius as a percentageof the production propeller radius. As the radius increases theefficiency increases and both the scund level and A-weightedlevel increase. The noise is changed by altering the tip

Report No. 4764Bolt Beranek and Newman Inc.speed. Be~ause the rotation is fixed at 2700 RPM,is directly proportional to the propeller radius.the tip speedThe two noise reduction strategies were combined asfollows: A min:mum induced loss propeller was designed with thecharacteristics of the production propeller listed in Table2.Next, a second propeller was designed with three blades. Thischange reduced the sound level anci increased the efficiency.Finally, the radius of the three blade propeller was reduced to90% of the original value. This further reduced the noise andreduced the efficiency down to its original value. An off-designanalysis shcwed that this propeller would operate approximatelythe same as the original. This meant that it would be quieterthan the original propel. er for all flight velocities. Its tipspeed would always be lower, because of the smaller radius, andthe source strength would always be 2/3 of that of a comparabletwo blade propeller.Lhta ReviewOSU has flight tested the low noise propeller. Aerodynamicobservations indicate no degradation in performance at the designpoint. That is, at full throttle during level flight, the enginerotated at 2700 RPM and the aircraft flew at the e-.me airspeed.Acoustic comparison of the new and old propellers is givenin Fig. 5. The horizontal axis in flight speed and the verticalaxis is the sound level measured with a wing-mountedmicrophone. This microphone was mounted in the disk plwe at1.90 (6.22 ft) from the propeller axis. The measurements weretaken at several. airspeeds during level flight at 1524M (5000 ft)altitude. At all flight speeds the new propeller is four to six

Report No. 4764Bolt Beranek and Newman Inc.signatures for tt: "Ygh speed flight. Again measurement andpredictions agree. However, some systematic differences areapparent. The predicted negative amplitude exceeds the measureeamplitude for the first and fourth pulse in Fig. 8. Thisindicates that one blade emits slightly less noise than theothers. Errors in the blade angle setting were recorded by OSU.Fence, the small negative sour3 pulse is most likely due to asmaller blade load.ConclusionsFlight tests on a new low noise propeller have demonstratedthat this propeller is 4 to 6 dB quieter than the productionpropeller with no measureable reduction in aerodynamicperformance. This propeller was designed using proceduresdeveloped at MIT during a study of the noise and performan~e ofpropeilers for light aircraft.A comparison of the calculated acoustic signature to themeasured signature shows that the theory accurately describes thenoise from the new propeller. This study demonstrates that; thetechnology is now available for the control of noise in generalaviation propellers.

Report No. 4764Bolt Beranek and Newman Inc.o~!S~~;gpL. 1sOF POOR QUALinAERODYNAMIC DESIGN CONSIDERATIONSENGINE PROPELLER /-POWER NEEDED TOOVERCOME AIRCRAFTDRAG-cc 30 40 50 60 70VELOCITY (midFIGURE 1.

Report No. 4764Bolt Beranek and Newman Inc.OR\QNC?- ?AGE ISOF pOOR QuAL~~SYNTHESIS OF ENGINEPROPELLER CHARACTERISTICSRPMFIGURE 2.

Report No. 4764Bolt Beranek and Newman Inc.QW!Gi!CAp $AGZ ioOF POOR Q: JALIWPROPELLER DIAMETER EFFECTS4 4110 100 120PROPELLER RADIUS% -FIGURE 3.

Report No. 4764Bolt Beranek and Newman Inc.0 ~ ~ i HPAGENOF POOR QuAU~BLADE NUMBER EFFECTS-2 4 6 8 10BLADESFIGURE 4

Report No. 4764Bo1 t Beranek and Newman Inc.IFIGURE 5,

Report No. 4764Bolt Beranek and Newman Inc.MEASURED AND PREDICTED SOUNDPRESSURE LEVELSMEASURED-70 80 90 100 110 117INDICATED AIR SPEEDFIGURE 6

Report No. 4764Bolt Beranek and Newman Inc.W>130MEASURED AND PREDICTEDFOURIER SPECTRArlto -f ' 'Wa1W 9 0 -E0t70-2dB[piiGqMEASURED0O-#T-Q0--1 2 3 4 5 6 7BLADE PASSING HARMONIC --,FIGURE 7

Report No. 4764Boly Beranek and Newman Inc.MEASURED AND PREDICTEDPRESSURE SIGNATUREFIGURE 8

Report No. 4764BGlt Beranek and Newman Inc.APPENDIX AFURTHER COMPARISONS OF MEASUREMENT TO THEORYOhio State has summarized their acoustic observations fortwo flights. On flight 30, the MIT propeller was set at approximately0.2O less than its nominal pitch; on flight 34, the MITpropeller was set at approximately 2.0° less than its nominalpitch. During each flight a number of runs were made. Each runcorresponds to a particular aircraft velocity. A summary offlight conditions is given in Tables A1 and AZ.Information on the sound level Is also given in thesetables. The column labeled "9verall SPLw contains theinformation supplied by OSU. The column : .oeled "MeasuredPropeller SPLw is computed at BBN. The height of each propellerharnonic amplitude given in the Fourier spectra is measured witha caliper. The harmonic amplitudes ar13 then addedlogarithmically to give the measured SPL. The column labeled"Calculated Propeller SPLw contains the predizted sound levelscomputed at BBN.The overall SPL exceeds the measured propeller SPL in TableA 1 and A2 because of the influence of background noise. There isone significant exception to this rule. These are for flight 34,run 134 rule. In this instance the measured propeller SPLexceeds the overall SPL. This is probably due to plottererror. This is also the only run in which the measured spectraexceeds the calculated spectra by a significant amount.Figure A1 is an example of the measured Fourier apectra.Superimposed on this figure is a line indicating the engine

Report No. 4764Bolt Beranek and Newman Inc.harmonics. The propel~sr harmonics are labeled Pi. The MITpropeller has thcec blades, hence the fundamental blade passingfrequency is three time8 the shaft frequency. The engine hasfour cylinders, each cylinder fires at one half the shaftfrequency, hence the fundamental engine frequency is twice theshaft frequency. The engine and propeller harmonics areidentical at every other propeller harmonic starting with P2. Ofparticular interest is a second set of harmonics labeled Ii.Each of these is at the propeller frequency minus the shaftfrequency. Not all of these harmonics correspond to engineharmonics. In particular I2 and I,, are neither propeller norengine harmonics. That the second, fifth and ninth engineharmonics are buried in the noise indicates that the sequence ofpeaks labeled Ii (for interaction harmonic) constitute a separatephenomena. One possible cause of these interaction harmonics isa non linear interaction between the sound from the engine andthe sound from the propeller. This interaction can yieldadditional tones at sums and differences of the engine andpropeller frequencies.Figures A2 to A17 compare the measured xnd predictedacoustic spectra. Closc agreement is evident in all cases exceptfor run 134 (Fig. A17). The predictions were made given only theshape and motion of the propeller. First the blade loads werecalculated, then the acoustic signature was calculated. Noempirical adjustments are needed.

Measured CalculatedRun Velocity Rotation Aircraft Overall Propeller PropellerRate Weight SPL SPL SPLReport No, 4764 Bolt Beranek and Newman Inc,FLIGHT 30AIR DENSITY 1.065 kg/~3SOUND SPEED 333 m/sALTITUDE 1524MRPMTABLE A1

Report No. 4764Bolt Beranek and Newman Inc.FLIGHT 34AIR DENSITY 1.076 kg/~3SOUND SPEED 333 m/sALTITUDE 1524 MRun Velocity RotationRateRPM277926692548Aircraf tWeightNt108001079110782OverallSPLMeasuredPropellerS" LCalculatedPropellerSPL253224491077310764235223162265107551074610737TABLE A2

Report No. 4764Bolt Beranek and Newman Inc.FIGURE A-1

IReport No. 4764Bolt Beranek and Newman Inc.ORIGih'AS. FASE 18OF POOR QUALITYFIGURE A-2

Report No. 4764 p~;c !PJ Bolt Beranek and Newman Inc.Gx,~\l:&I A L ~~7 p03R Qb00oI]12- 18-83RUN NO. 120FUNO FRE@= i 27.3HZa SPL= i 15- 9 DR(MEASURED)PJ@ THEORYc..5 AHERTZ BANDWIDTH0 200- II I I I 1400 600 800 1000FREQ HZ 4 0FIGURE A-3

Report No. 4764ORIGINAL PACE rsOF POOR QUALITYBolt Beranek and Newman Inc.P12- 15-50RUNNO. 121FLIND FSEQ= i26.-2HZSPL= 115.8 DB(flEASURED)Q THEORY,5 HERTZ BANDWIDTH0IFIGURE P-4A-8

Report No. 4764(-Jz;;;:,'..i p;,-;zOf PODR QUALITYBolt Beranek and Nman Inc.0 THEORY,5 HERTZ BANDWIDTHFIGURE A-5

Report No. 4764Bolt Beranek and Newman Inc.Pi3 -a:- I0 i i- 1!Itf:C!i? FFiFG= i 17. 3tiZ';PL- i13.6 [!R (MEASURED)Q THEORY,5 HERTZ BANDWIDS'HFIGURE A-6

Report No. 1764Bolt Beranek and Newman Inc.ORIGfRilL PA22 :3OF POOR QUALITY12- 18-80RLil4 NO. 124FUI.JD FREQ= i 13.OHZn SI'Lz 112.2 ni; (MEASURED)o0I0 THEORY,5 HERTZ BANDWIDTHFIGURE A-7

Report No. 4764Bolt Beranek and Newrnan Inc.00C) 12- 18-80NRUN I\JC1. 125FUND FREQ= i12.9HZSPLZ i11.6 I38 (MEASURED)0 = THEORYO, j = HERTZ BAE!DHIDTHFIGURE A-8

Report No. 4764ORIG!~J:L p3 c?- . -. - .- i.>OF P03& C QRL~~Bolt Beranek and Newman Inc.00o 12- 18-80- RUN NO. 126FUND FSEQ= il4.1HZSPL= 1 12. 2 DB(f"lEASURED)o0 0IQTHEORY,5 HERTZ BANDWIDTH'1 IOr------ I I v-1rd 0 200 4G0 600, 800 1000FREQHZ "10FIGURE A-9

Report No. 4764Bolt Beranek and Newman Inc12- 13-80Rut4 NO. 127FUt\!E FREO= i 39- OfiZSPL= i19.1 DB (MEASURED)0 THEORY,5 HERTZ BANDWIDTH7 9200 400 600FREQHZ n1U'FIGURE 10

Report No. 4764Bolt Beranek and Newman Inc.ORIGENAL P.'.!: *: :;OF POOR QUALITY9 I0 THEORYiQ ,5 HERTZ BANDYIDTHIIQFIGURE A-11

Report No. 4764Bolt Beranek and Newman Inc.000 m12- 18-80-1RUNNCI. 129FUND FSEGI= i26: 3HZSPL= i 15. S DB (MEASURED)3 0 THEORYFIGURE A-1 2

.. ,Report No. 4764Bolt Beranek and Newman Inc.ORIGENFaL PAGE tSOF POOR QUALIM0 -03 12-13-80N,- RUN 140. 130FUND FREE; 126.5MZa SPL= ilS.4 DB (MEASURED)clo THEORY'33A,5 HERTZ BANDWIDTHC[L -LD0a3,PFI GURE A-1 3

Report No. 4764Bolt Beranek and Newman Inc.L1n:HUN I 13i-1IFUIIC F9EQ- 121-OHZo 1 SPL= 113.7 DB (MEASURED)J 1. . 0 THEORY,5 HERTZ BANDWIDTHFIGURE A-14

Report No. 4764 B )l t Beranek and Newman Inc.ORIGINAL PAGE ISOF POOR QUALITY0C)3 12- 18-80RCiiL' NO. 132FUNG FRE@= ilS.8tiZSPL- i12.6 TJR(MEASURED)='!Q THEORY,5 HERTZ BANDWIDTH0 IFIGURE A-15

Report No. 4764ORIGINAS. FAG5 !3OF POOR QL:AL!T'IPol t Beranek and Nemn Inc.12- 18-50RUN NO. 133FU~\!D FREE= i 1 3 . 7 ~ ~SPL=@ TH EOgY85 . HERTZ BANDIYIDTHill-8 Gi3(MEASURED)F:RE@ HZ 4 C J iFIGURE A-16

Report No. 4764 Bol t Beranek and Newman I nc .ORIGINAL PAGE ISOF POOR QUALlrV12- 17-80Fitif\! biz. 134FUKD FSE@=SPL=i i3.2HZ112.1 DB(MEASURED10 THEORY,5 HERTZ BANDWIDTHFIGURE A-1 7

Report No. 4765Bolt Beranak and Newmnn Inc.APPENLIIX BNOTES ON TYE DATAOhio State has spent some tine developing a data package tosummarize the fllght test measurements. For each run a packet ofinformation is supplied, This packet consists of two tables andthree graphs. There ape some small inconsistencies in thisinfomation package which are reviewed here.Tables B1 and 82 are supplied by OSU for each run. Table 81summarizes the engice operating conditions. Table B2 s*mmarizesthe frequency spectra. On Table B1 there are several propellerrotation rates indicated: 2700, 2541, and 2697 RPM. The firstrate is the approximate reading of the flight engineer, ?.hesecond rate is the tachometer reading and the last rate AS thatdeduced from the Fourier spectra. In all the BBN calculations,only the rate deduced from the Fourier spectra is used.Table B2 contains two sets of measurements. One set givesthe cnwputed harmonic amplitudes using an FFT routine, the secondset gives the results of a real time spectrum analyzer. Theresults of the FFT do not agree with the spectrum analyzer.Furthermore, the tabulated data appears to be the peak amplitudeat each harmonic and not the RMS amplitude.Let A be the amplltude of the ith harmonic given in Table02. If the sound Icvel 1s 20 log (Ai/Z r loo5), then this soundlevel exceeds that given in the plotted spectra hy approximately3 dB. Because 3 dB is approximately 20 log 2, the amplitudesliated In Table B2 are most probably the peak amplitudes at each

Report No, 4764Bolt Berurek and Wenun Ino,Three figures are also su?plied with each run. Fig, B1 is aplot of the peak sound level at each harmonic. This 18 a plot ofthe column labeled FASTFT in Table 82. Fig. 82 is a plot of theaound level versus frequency, Each of the peaks in Fig. B2labeled P1 is a propeller harmonic. These peaks are 20 logA 2 x log5) where Ai is the amplitude listed in Table B2.Fig. 83 presents the pressure time signature.There appears to be two errors in the time signature.First, if the period is measured with calipers then the measuredperiod is .0235f.002 aec, which corresponds to a rotation rate ofapproximately 254 1 RPM. However, the measured blade passingfrequency in Fig. B3 is 13624 Hz which corresponds to a rotationrate of approximately 2697 RPM, This is a 6% error. Apparentlythe pressure time signatures are based on the tachometer RPM.Because the tachometer reading is saspected to be in error, theperiod should be corrected as is done in Fig. 8 of this test.The second error in the nteasured pressure time signature isthe sign of the amplitude. The pressure signature in Fig. 83shows a slow drop in pressure followed by a steep rise. Also,the peak negative amplitude is larger than the peak positiveamplitude. The calculated pressure signature shows precisely theopposite effect. In the calculation a slow rise in pressure isfollowed by a steep drop. Also the peak positive emplitude islarger than tbe peak negative amplitude. The error is one ofsign, and the data can be corrected by multiplying all measurementsby -1. The probably source of error is the interpretationof the voltage reading from the microphone. For condensor microphonesa negative voltage corresponds to a po8itive pressure.This correction 1s also made on Fig. 8 in this test.

Report No. 4764ORLCONAL PAGE CS,;,.I ,,- Lj,;fr: ,..ITPC)(~IC) l:.~ nl-~r ~LA?.JF. FULL ~WUTT~_F -( 193=117 YPH), ".,\ ;I [J r- 2:CLLIaQbTION DATA UCFD:.- --- ---- - . --- - ----. - ---- -- .- - - - -- - . -. - -+ i f SLrroF (IJW~TC/VOLT~ ZtRO (VOLTS) REF VALdE- *:-$ 1 - - - --. --qA?k!FOLD P?ESSU?E -9,141730E+02 o,ooZ-- 0~293€+02Ov.YAMI C PQESSUQE -n,4957Q!IE+nO -0.032-0,00OE+01lh6LE OF CTTPCK -0 &QCO~OE+OZ 0.000E+01 . . oTGl. AI ---- ------- . ----- ----- -0- - - - --. ----.- 3 L43ACCEL n,\35nco~q11 0.039: f f O ~ ~ A Q Gpay n. v~nn0~+06 -0.0 04 O.CJOOE+~~. . - ..'RkG PATA *LOT-.- - -- .- . - - -- - -- . --- -- . --\ LIN-LIN PLOT'-~f LIN-LO5 PLOT TC-;: f r .: t .[ ..C~LC'JLLTE~. .- FU~J~)P*AEN! 4~ F~FCI~FNCY IS 139.9 ~7 C~LC ;IPm= e657-NiTw art GUS ~t 1:-.(F PC~XO!') OF 0.25)il'C)it:~~-j S~;CO~J::SNr S!JiTIt.iG SOUbifr PD:S~?IQF 1-FVEL IS 11t5.4 OH . - r---.- - -- - - -- .--- .TABLE B-1

Report No. 4764ORIGINAL PAGE ISOF poon g u mBolt Beranek and Netman Inc.- u;(L) - =--.- 11~~35.-~ -- B---- W ) = 107.41, Dace) = iis.3z. D%C) = 118-19.- --- -.-- -.-.---- - .-- -- .TABLE 8-2

Report No. 4764ORIGINAL PAGE iSOF POOR QUALIKBolt Beranek and Nernnan Inc.= HERTZ BANDW I DTHFIGURE 0-1

Report No. 4764Bolt Beranek and Newman Inc.ORIGtNAL PAGE iSOF POOR QUALTPr:ISPL- =.5 = HERTZ BANDWIDTH. D (MEASURED)

Rewrt No. 4764Bolt Beranek and NemMn Inc.FIGURE 8-3

Report NO. 4764Bolt Beranek and Newvan Ine.Ref erencea1. Succi, et al. nNoise Performance of Propellers forLight Aircraftw NASA CR 165732, June 19812. Succi, et al. nExperimental Prediction of PropellerNoise Prediction." AIAA 80-0994, June 19803. Houghton and Brock. wAerodynamics fop EngineeringStudentsw. Ed Arnold (Pub), London 19664. "Federal Aviation Noise Standardsw. Title 14 Code ofFederal Regulations, Chapter 1, Part 36.5. Larabee, wPractical Design of Minimum Induced LossPropellersw. SAE 790585, April 1979.6. Succi, nDesign of Quiet Efficient Propellersn. SAE790584, April 1979.