Aero in Aeroelasticity

8/22/2019 Aero in Aeroelasticity

1/24

NASA/TM-2000-209589USAAMCOM-TR-00-A-005

Putting the Aero Back Into AeroelasticityWilliam G. Bousman

March 2000


2/24

The NASA STI Program Office... in Profile

Since its founding, NASA has been dedicated to theadvancement of aeronautics and space science. TheNASA Scientific and Technical Information (STI)Program Office plays a key part in helping NASAmaintain this important role.The NASA STI Program Office is operated byLangley Research Center, the Lead Center forNASA's scientific and technical information. TheNASA STI Program Office provides access to theNASA STI Database, the largest collection ofaeronautical and space science STI in the world.The Program Office is also NASA's institutionalmechanism for disseminating the results of itsresearch and development activities. These resultsare published by NASA in the NASA STI ReportSeries, which includes the following report types: TECHNICAL PUBLICATION. Reports ofcompleted research or a major significant phaseof research that present the results of NASAprograms and include extensive data or theoreti-cal analysis. Includes compilations of significantscientific and technical data and informationdeemed to be of continuing reference value.NASA's counterpart of peer-reviewed formalprofessional papers but has less stringentlimitations on manuscript length and extentof graphic presentations.TECHNICAL MEMORANDUM. Scientific andtechnical findings that are preliminary or ofspecialized interest, e.g., quick release reports,working papers, and bibliographies that containminimal annotation. Does not contain extensiveanalysis.CONTRACTOR REPORT. Scientific andtechnical findings by NASA-sponsoredcontractors and grantees.

CONFERENCE PUBLICATION. Collectedpapers from scientific and technical confer-ences, symposia, seminars, or other meetingssponsored or cosponsored by NASA.SPECIAL PUBLICATION. Scientific, technical,or historical information from NASA programs,projects, and missions, often concerned withsubjects having substantial public interest.TECHNICAL TRANSLATION. English-language translations of foreign scientific andtechnical material pertinent to NASA's mission.

Specialized services that complement the STIProgram Office's diverse offerings include creatingcustom thesauri, building customized databases,organizing and publishing research results.., evenproviding videos.For more information about the NASA STIProgram Office, see the following: Access the NASA STI Program Home Page at

http://www.sti.nasa, gov

E-mail your question via the Internet tohelp @ sti.nasa.gov

Fax your question to the NASA Access HelpDesk at (301) 621-0134

Telephone the NASA Access Help Desk at(301) 621-0390Write to:NASA Access Help DeskNASA Center for AeroSpace Information7121 Standard DriveHanover, MD 21076-1320


3/24

NASA/TM-2000-209589USAAMCOM-TR-00-A-005

Putting the Aero Back Into AeroelasticityWilliam G. BousmanAeroflightdynamics DirectorateU.S. Army Aviation and Missile CommandAmes Research Center, Moffett Field, California

National Aeronautics andSpace Administration

Ames Research CenterMoffett Field, California 94035-1000

March 2000


4/24

Available from:

NASA Center for AeroSpace Information7121 Standard DriveHanover, MD 21076-1320(301) 621-0390

National Technical Information Service5285 Port Royal RoadSpringfield, VA 22161

(703) 487-4650


5/24

Putting the Aero Back Into Aeroelasticity

William G. Bousman

ArmyNASA Rotorcraft DivisionAerofl ightdynamics Directorate (A VRDEC)US Army Aviation and Missile Command

Ames Research Center, Moffett Field, California

AbstractThe lack of progress in understanding the physics of

rotorcraft loads and vibration over the last 30 years isaddressed in this paper. As befits this extraordinarilydifficult problem, the reasons for the lack of progress arecomplicated and difficult to ascertain. It is proposed herethat the difficulty lies within at least three areas: 1) a lossof perspective as to what are the key factors in rotor loadsand vibration, 2) the overlooking of serious unsolvedproblems in the field, and 3) cultural barriers that impedeprogress. Some criteria are suggested for future researchto provide a more concentrated focus on the problem.

IntroductionThe dynamics community has necessarily focused on

the twin issues of rotor loads and vibration since thebeginning of the industry. There has been gradualprogress in recent decades in our ability to accommodateloads and to reduce the effects of vibration, but it is thepremise of this paper that our understanding of the physicsof the aerodynamic loading and response of a rotor has notshown a significant improvement in the last thirty or soyears. It was well known in the 1950s that a uniformwake was not suitable for the accurate prediction of rotoraerodynamic loading. Piziali and Du Waldt (Ref. 1)demonstrated a substantial advance in prediction capabilityby incorporating a prescribed wake in their analysis.Landgrebe (Ref. 2) showed that it was possible to modelthe wake as a distorted or free wake. By the early 1970s itappeared that the representation of the basic physics of therotor was in place. The 1973 AGARD "SpecialistsMeeting on Helicopter Rotor Loads Prediction Methods"in Milan in some ways represents a watershed in thedevelopment of analytical methods. Loewy (Ref. 3) sawthe progress that had been made when he said, "Instead of

running into unexpectedly high loads almosteverywhere the first time the full flight envelope isexplored, we now only run into them occasionally, atsome extreme flight condition." Piziali (Ref. 4), in hiscommentary on the conference, was less sanguine andexpressed his opinion that the advances of the past decadehad been in the scope of the analyses, but not in theiraccuracy. He felt that the structural problem was well inhand, but that further advances would requireimprovements in the aerodynamic models.

The past decades since the Milan meeting have beenlargely marked by incremental improvements in theprediction of rotor loading, although there are few caseswhere these "improvements" have been demonstratedclearly and systematically. In the area of rotor loadscalculation the critical design loads that are computed inmaneuvers are sometimes of the correct size, but theysometimes have the incorrect phase, as shown by Sopherand Duh (Ref. 5) for a Sea Hawk. In one sense theseloads are suitable for design, but the phase errors show adeficiency in the modeling of the physics that offers acautionary note. For vibration, the present comparison ofanalytical methods with flight test data is not satisfactoryas recently shown by Hansford and Vorvald (Ref. 6). Inparticular, they conclude that the structural modelingcapability represented by the recently-developed finiteelement methods shows no advancement in predictiveaccuracy over classical modal techniques. The problem ofprediction has perhaps been best characterized by Johnson(Ref. 7) who said, "For a good prediction of loads it isnecessary to do everything right, all of the time. Withcurrent technology it is possible to do some of the thingsright, some of the time."

The purpose of the present paper is threefold. First, Iwill attempt to put the loads and vibration problems intheir proper perspective and at the same time attempt to


6/24

demythologizeomenfortunaterendsfthepastdecade.Second,willoffersomeampleroblemsnloadsandvibrationhatarepresentlynsolvednd willsuggestthatrealprogressustbemadentheseroblemseforetheresanyhopeofadvancinghestateof theart.Third,I willdiscussnumberfculturalproblemselatedoprogressnthesciencesndoffersomemodestsuggestionsorimprovementn thesereas.A Search for Perspective

The problems of loads and vibration have always beena part of helicopter development and in this sense havebeen at the forefront of all efforts by dynamicists in theindustry. By necessity, the primary motivation hasalways been to solve or reduce the problems that crop upduring the helicopter' s development phase. If at the sametime an advance in our understanding of the basic physicsis achieved then this is well and good, but thatunderstanding has never been the primary objective. Inthe last 20 years the contributions of academia to theseproblems have become more important and in a sense theacademic perspective is reversed from the industrialperspective. That is, the academic practitioner needs tounderstand the physics, even if this does not leadimmediately to a practical result, while the industrialpractitioner must achieve a practical result even if thephysics is not fully understood. The best of all possibleworlds is when a balance is achieved between theseconflicting objectives, but sometimes things go awry. Insome sense the dynamics community has lost itsperspective on a number of topics in recent years and Iwill offer my view on what is needed to return to center inthree areas. First, in our attempts to find the "silverbullet" that will solve all our problems we tend tooversimplify the task at hand and suggest that there isreally only one best approach. I believe this approachwill always be fruitless and I attempt to illustrate thisfruitlessness with an example. Second, the physics ofvibration are enormously complex and we need tounderstand the differences as well as the similarities indifferent flight regimes. Third, descending flight, which isso important for blade-vortex interaction (BVI) noise, is abenign environment for vibration and new vibrationcontrol approaches should not be demonstrated for theseconditions but, rather, for the low-speed and high-speedvibration regimes.Which Is the Silver Bullet-Rotor Wake orBlade Elasticity?

When analytical methods become sufficientlydeveloped so that they can consistently and accuratelypredict the loads and vibration for a new rotorcraft, then it

will be possible to determine the relative importance ofvarious aspects of the modeling problem. Until that timecomes we are reduced to endless arguments over whetherthe rotor wake is more important than blade aeroelasticity,or whether torsional deformations are more important thanbending deformations, or whether the trim solution is thekey to the problem.

In 1990 a sample problem of a Puma rotor in high-speed flight was examined using a variety of analyticalmethods in a stepwise fashion (Ref. 8). Ten calculationswere established, as shown in Table 1, p rogressing fromthe simplest possible representation of a rotor, to the fullcalculation with all capabilities of the analysis exercised.The first calculation, Case 1, used a uniform wake (noharmonic variation), linear airfoil section properties, nounsteady aerodynamics, no radial flow corrections, and theblade dynamics were represented by a rigid, hinged blade.The rotor configuration was made progressively morecomplex in a stepwise fashion. For Case 3, a prescribedwake model replaced the uniform wake. For Case 5,nonlinear airfoil properties were used, based on a set ofequations that represent the NACA 0012 airfoil, instead ofthe linear aerodynamic properties. For Case 7, unsteadyaerodynamic terms were added to the model. Radial flowcorrections were added for Case 8, and in Case 9 flap andlead-lag bending modes were incorporated in the model.Case 10 added the blade torsion modes and represented an"all-up" calculation. Note that three cases were includedwhere the root cutout was increased from 0.228R, thevalue for the Puma, to a value that would exclude reversedflow for the sample problem (Cases 2, 4, and 6).

The high-speed case examined was based on a flighttest condition for the research Puma with an advance ratio,It, of 0.38; thrust coefficient over solidity, CT/O, of 0.08;shaft angle relative to the freestream, a s, of -6.8 deg; firstharmonic cosine flapping, 131c, of 0.39 deg; and firstharmonic sine flapping, _ls, of-0.07 deg. However, thesample problem was for a rectangular blade, rather thanthe swept-tip blade of the research Puma. Four analyticalmethods were used to examine this problem. TheWestland/DERA code, developed collaboratively byWestland Helicopters and the Defence Evaluation andResearch Agency (DERA) in the United Kingdom, wasused by Colin Young of DERA. The R85/METAR code,developed by Eurocopter France, was run by FrancoisToulmay of Eurocopter. The CAMRAD/JA analysis wasrun by Thomas Maier of the AeroflightdynamicsDirectorate, while the original CAMRAD analysis wasused by Neil Gilbert of the Aeronautical and MaritimeResearch Laboratory (AMRL) in Australia.

The comparison of these calculations in Ref. 8indicated that the effects of reversed flow, nonlinear and


7/24

Table 1. - Stepwise analytical model for high-speed rotor case. rc is the radius of the root cutout and R is the rotor radius.

Case rc/R Wake Aero Unsteady Radial Bending TorsionTables Aero Flow Modes Modes

l 0.228 uniform constant no no no no2 0.400 uniform constant no no no no3 0.228 prescribed constant no no no no4 0.400 prescribed constant no no no no5 0.228 prescribed NACA 0012 no no no no6 0.400 prescribed NACA 0012 no no no no7 0.228 prescribed NACA 0012 yes no no no8 0.228 prescribed NACA 0012 yes yes no no9 0.228 prescribed NACA 0012 yes yes yes no10 0.228 prescribed NACA 0012 yes yes yes yes

unsteady aerodynamics, and radial flow had little effect onthe blade vibratory loads. The important effects, by andlarge, were seen in comparing Cases 1, 3, 9, and 10. Thecalculations for vibratory normal force at 0.95R are shownin Figure l for the four calculations. There is no Case 10calculation for R85/METAR as the analysis would notconverge with the torsion degree of freedom added. Thevibratory airloads are largely 3/rev because of the reducedor negative lift that occurs at the beginning of the secondquadrant. Generally, the effect of the prescribed wake(Case 1 to 3) was to reduce the amplitude of the 3/revloading, but increase the amplitudes of higher harmonics.The addition of bending modes (Case 9) and torsion modes(Case 10) tended to increase the harmonic loading.

The differences between the cases (or between thecodes) for normal force can be assessed quantitativelyusing harmonic correlation (Ref. 9). In this approach onecase is considered the reference condition or independentvariable and the second case is considered the dependentcondition. The sine and cosine harmonic coefficients ofthe dependent case are then plotted against the harmoniccoefficients of the reference case for any range ofharmonics. A least squares line is fitted to the resultingplot and the slope, m, and correlation coefficient, r, arecomputed. A slope close to one indicates very goodagreement between the reference and dependent cases, and acorrelation coefficient near one means there is very littlescatter. This is illustrated in Figure 2 for theWestland/DERA calculations for four combinations. Thefirst subfigure shows Case 1 as a function of Case 3. Thesolid circles represent the individual sine and cosineharmonics and the solid line is the least squares fit. Theslope and correlation coefficient are indicated on the figure.Perfect correlation is indicated by the dashed line. For thiscomparison the calculated slope and correlation coefficientare largely dependent upon the 3rd and 4th harmonics of

the lift. The effect of the prescribed wake for thiscalculation is to reduce the size of these lower harmonics,but to increase the size of the higher harmonics. In thesecond subfigure, Case 3 is compared to Case 9, that is,the effects of adding blade bending modes are examined.

In this case more scatter is seen in the comparisonand the addition of elasticity has a significant effect on thephase of the harmonics. There is also, on average, anincrease in the amplitudes. The additions of the torsionmodes, Case 10, is illustrated in the third subfigure. Thescatter is reduced in this case and there is also a reductionin the harmonic amplitudes. The final subfigure comparesthe full-up calculation, Case 10, with Case 1 and there isalmost no correlation between the two calculations. Thisfinal result is not surprising, of course.

The differences in the stepwise calculations for thefour analyses is summarized in Table 2 using theslope/correlation coefficient pairs for each comparison. Itis interesting to note substantial difference between thesemethods for the various steps. The Westland/DERA,R85/METAR, and CAMRAD/JA codes all show a generalreduction in the amplitude of the vibratory loads when theprescribed wake (Case 3) is added, but CAMRAD showsan increase in these loads. Similarly, when blade bendingmodes are included (Case 9), the vibratory loadingincreases for the Wesfland/DERA code, but decreases forthe other three.

A cursory examination of Figure I suggests that eachof these analyses is getting about the same vibratoryloading. Yet when the calculations are examined in detail,it is seen that each analysis is showing differing effects ofthe wake, elasticity, and, perhaps, trim. In one sense thegenerally good qualitative agreement that is seen in Figure1 represents today's state of the art-our best analysesprovide reasonable results. But the comparison in Table 2


8/24

Case I (Glauert wake)Case 3 (prescribed wake)Case 9 (bending modes)Case 10 (torsion modes)

0.050.04

rZ 0.030.020.010.00

r_,_ -0.0!_n* -0.020 -0.03Z

-0.04-0.05

RB5/METARWestland/DERA 0.000.04oo_ 0.02

-_- . " . _ 0.00V g -0.020 --0.03 ;/ z -0.04

I I I i I i I --0,05oo 13es too ZZ5 zvo 3_5 3soBLADE AZIMUTH, deg

I I I I I I I Ii I45 0 45 80 135 180 225 270 315 300BLADE AZIMUTH, deg

0.05 CAMRAD/JA 0.05 CAMRAD0.04 0.04

r_ 0.03 r,.) 0.03oo..,,, /, ooo, /,.._oo,...."fll-,V -.-.o, r--_i _,.C - "


9/24

,.J

t--:r._,-1ZOp.t..9CO

Case I compared0.03

0.02

0.01 4s

0.00 I I 6 I'

-0.01 ,_ss* **s-0.02 . SS

P I-0.03_y03_0.021 -0.01 0.00 0.0!

SECTION LIFT. M2CL

/to Case 3 //3C e/ ss* 0.03

,_,s,, SSS*S 0.023s . rj

**e_..41* _ 0.01-:i i i ,,_I 0.00

Zm ffi 1.204 0r= 0.888 _ -0.01U

-0.02

i i -0.030.02 0.03 -,_

0.03

0.02

0.01

_-" 0.00

-o.oi

-0.02

-0.03 i I I-0.03-0.02 -0.01 0.00 0.01 0.02 0.03

SECTION LIFT, M2CL

Case 9 compared to Case I00.03

ss S

s

s** 0.02

J 0.01

_, i , , _ 0.00

m = 0.645 OZr = 0.715 5 -0.01

sos

est I !

Case 3 compared t.o Case 9%/... 3C/**C .s***

4s _ 3sIT

' : -- ID I I I

, S*SSS# s _S ' 6S_

D3-0.02 -0.01 0.00

m = 1.135r = 0.474

I I I0.01 0.02 0.03

SECTION LIFT. M2CLCase I compared to Case 10

) 3c ssss

s Ss S

o Sss S

. s S

m: 0.285r 0.09?

-0.02 ss Ssss

-0.03 ' i i I i l-0.03-0.02 -0.01 0.00 0.01 0.02 0.03

SECTION LIFT, M2CLFigure 2. - Harmonic correlation showing stepwise comparisons between cases for Westland/DERA analysis; 3-12harmonics.

Table 2. - Slopecorrelation coefficient pairs from harmonic correlation for stepwise calculations; harmonics 3-12.

Westland/DERA R85/METAR CAMRAD/JA CAMRADCase 1/3 1.20/0.89 1.23/0.76 1.09/0.97 0.84/0.97Case 3/9 1.14/0.47 0.84/0.90 0.74/0.60 0.86/0.53Case 9/10 0.65/0.72 - 0.67/0.93 0.58/0.96Case 1/10 0.29/0.10 0.88/0.58(1) 0.18/0.18 0.27/0.31

(1) Case 1/9.


10/24

thevibrationwithabsorbers,solators,ractivecontrols.However,hesehysicsecomemportantf featuresfthebladedesignretobeusedasameansfvibrationcontrol.Thebladeaerodynamicsndherotorresponseareverydifferentnthetworegimesndt isessentialokeepheseifferencesnmindnattemptingomodelandunderstandibratoryoads.Figurecomparesormalorcemeasurementst0.95Rwithcalculationsortwoaircraft.TheairloadmeasurementsortheUH-60Awereobtainedn1993(Ref.10)andarenterpolatedrommeasurementst0.92Rand0.965R,whilethecalculationsrefromRef.9. TheresearchumameasurementsndcalculationsrefromRef.9. ThecalculationseremadesingheCAMRAD/JAnalysisithbothprescribedakeandreewakemodels.t low speed the prescribed wake is unableto capture the rapid variation in aerodynamic loading, butthe free wake model obtains significantly better results.At high speed there is no difference between the two wakemodels. For a conventional rotor the importance of thefree wake declines as airspeed increases and little differenceis seen between the wake models for advance ratios above0.2 (Refs. 9, 11).

The importance of the free wake for vibrationprediction in low-speed flight has been illustrated in Ref.9 and figures from that study are repeated here. Figure 4shows both a blade plane view and a disk plane view for acalculation for the research Puma at kt = 0.098 for theprescribed wake calculation. The blade tip vortices appearas a sequence of epicycloids that overlap as seen in thedisk plane view from an azimuth of about 60 to 85 deg.If the tip vortices are cut in the _ = 75 deg plane, asshown in Figure 4a), it can be seen that the vortices arestacked sequentially (prescribed), one after the other. Theeffect of each tip vortex, therefore, is lessened as thevortex wake is convected beneath the plane of the rotorand the combined effects of these well-spaced tip vorticesdo not dominate the airloads. In the case of the free wake,as shown in Figure 5, the wake epicycloids appear muchthe same in the second quadrant, but in the first quadrantthere is considerable distortion as the various tip vorticeswrap around each other. The cutting plane at _ = 75 degshows that the nearest vortices to the blade tip are derivedfrom the -3 and -4 blade passages, where the notation -3refers to the tip vortex generated by the blade that passedroughly 270 deg prior to the sketch shown here. Notethat the -1, -2, -5 blade vortices are not nearly as close tothe blade for this cutting plane. The combined effect ofthese closely intertwined vortices creates the rapid airloadvariation that is observed in the first quadrant in Figure 3.A similar phenomenon is seen on the retreating side of therotor. Thus, computation suggests that the dominantvibratory loading source at low speed is the intertwining

of the rotor tip vortices that occurs near the outer edges ofthe rotor disk and that these pressure fluctuations are ofthe proper frequency content to cause the 4/rev vibratoryloads.

At high speed the situation is different. Theepicycloids that describe the tip vortex trajectories arespaced further apart and, as the rotor disk plane is tiltedmore forward, the tip vortices are convected more quicklyaway from the disk plane. The source of vibratory loadingon the blade at high speed is the impulsive-like negativeloading on the advancing side of the rotor that is seen inFigure 3. There may be secondary effects from the -1blade tip vortex, particularly when a vortex of oppositesign is created over the area of negative loading as seen onthe Black Hawk. Tung et al. (Ref. 12) have hypothesizedthat the small disturbance seen in the UH-60A airloads atabout _ = 90 is a consequence of a reversed rotationvortex. Nonetheless, the overall effect on the vibratoryloading at high speed of these secondary vortices appearssmall and it is essential to focus on the primary cause ofhigh-speed vibration which is negative loading in the firstand second quadrants.

A contrary view on the necessity of a free wake modelat high speed is provided by Refs. 6 and 13. Hansford andVorvald (Ref. 6) state that a free wake is required for theprediction of high-speed vibration, yet do not make directcomparisons between prescribed- and free-wake models.Wang (Ref. 13) also states that a free wake is necessaryfor a correct calculation of the vibratory airloads based onhis work with the UMARC/S analysis. Although hespeculates as to the cause of the differences, he does notprovide any airload comparisons for the prescribed- andfree-wake models. The importance of the free wake athigh speed will only be understood when all theresearchers involved approach this problem as one wherehypotheses are either affirmed or refuted within the normalconstruct of the scientific method.Low-Speed Vibration and BVI Noise AreDifferent Beasts

The very important problem of noise radiated fromblade-vortex interactions (BVI) in descending flight hasbeen studied extensively in recent years. Higher HarmonicControl (HHC) actuators have been used in wind tunneltests of a model rotor (Ref. 14, 15) to demonstrate thesubstantial reductions in the BVI noise that are possiblefor these conditions. However, in some cases thevibration is increased for the controls that reduce noise.Tests have also been performed on full-scale helicoptersand tiltrotors in a wind tunnel with Individual BladeControl (IBC). It has been shown that noise and vibrationcan both be reduced simultaneously with the addition of


11/24

Flight TestPrescribed WakeFree Wake

0.25

0.20

0.15

0.10O) 0.05

0.00re -0,05O'7 -0.10

-0.150

0.25

0.20

0.15

0.I00.05

r_0.00

g -0.05

-0.10

-0.15 0

UH-6OA, # = 0.095 0.25 Puma,/z = 0.0980.20

__) 0.10_ 0.05

" _ 0.00re -0.050Z

-0,10

I I I I I ! I ! --0.15 I I I I I I I I45 0o 135 leo 22._ zvo 315 360 ,,15 00 135 leo 225 2';.0 315 300BLADE AZIMUTH. deg BLADE AZIMUTH. deg

UH-6OA. # = 0.355 0.25 Puma. p. = 0.3620.20

0.15

. _ 0.I0

: i , , i _ 0.00g -0.05

k_/ o -o,o' , , , , , , , , , , , , , , ,40 0o lSS 10o 220 2vo 310 .qeo 40 9o 130 100 220 2vo ._is 30o

BLADE AZIMUTH, deg BLADE AZIMUTH, deg

Figure 3. - Comparison of measured airloads for two aircraft and two flight speeds with predictions from CAMRAD/JA; r/R=0.95.

2/rev control for a four-bladed helicopter rotor (Ref. 16,17). Considerable effort has been taken to define theairspeed and descent conditions that result in the greatestBVI noise and it is these conditions that have beenexamined in the test programs. However, comparableefforts have not been made to identify maximum vibrationconditions and, based on UH_50A flight tests, it appearsthat the maximum noise conditions encountered in flighttend to be the most benign in terms of vibration. Thisresult suggests that these conditions are fundamentally notappropriate for testing closed-loop active controllers forvibration.

Crews has defined an Intrusion Index to represent howthe pilot or crew respond to helicopter vibration (Ref. 18,19). The Intrusion Index requires vibration measurementsin three orthogonal axes for each location in the helicopterwhere the index is calculated. The three axes are weighteddifferently: the lateral or y-axis has an 0.75 weight relativeto the vertical or z-axis and the longitudinal or x-axis hasan 0.50 weight. In addition, each component includes aweighting factor that depends upon frequency. TheIntrusion Index is the norm of the four largest harmonicamplitudes in each axis and, hence, is composed of 12harmonic amplitudes. Although the l/rev amplitude isexcluded, all other harmonics up to 60 Hz are included and


12/24

.2b

-_o

,'1

..4D.?Y

9 EBS

-_ 4)1

-6

-1

_r

-4@-5

I I : I I I.1t_ .Idt 1i1'_

_,1=1

-2-3

a)I I

1.I.


13/24

3.0

2.5

x 2.0

o

t.e-

1.5

1.0

0.5

...o.., Cw/_ = 0.08--I! b Crews (Ref. 18)

0.0 I I , I I0.0 0.I 0.2 0,3 0.4

#Figure 6. - Intrusion Index at pilot floor for UH-60A forlevel flight sweep at Cw/(r = 0.08 compared with Ref. 18.

It is expected that the norm of the Intrusion Indexshould be dominated by bN/rev harmonics and this is truefor an index computed in the transition range. Anexamination of the norm at I.t = 0.09 shows that thecombined effect of 4/rev and 8/rev accounts for 91% of theindex. However, at the maximum speed, _ = 0.37, the4/rev harmonics account for only 67% of the index. Theaddition of the 2/rev vertical brings the norm to 86% and6/rev vertical brings it to 91% and this demonstrates theimportance of non-bN/rev harmonics for some flightconditions.

During the UH-60A Airloads program (Ref. 10) anextensive set of data was obtained using ground-acousticmeasurements in cooperation with Langley ResearchCenter (Ref. 20). These data included level flight, climbs,and descents relative to a microphone array installed on theground. The Intrusion Index for the level flight cases thatwere flown during the ground-acoustic tests are comparedin Figure 7 to the Cw/c = 0.08 airspeed sweep data fromFigure 6. For the ground acoustic tests, the aircraft wasoperated at a lower weight than for the airspeed sweep dataand Cw/O was about 0.065. As seen in Figure 7, theIntrusion Index is about 0.5 units higher for the ground-acoustic tests and the source of the increase is not known.Most of these data were obtained between advance ratios of0.10 or 0.15 and 0.30 and this range of test conditionsexcludes the transition and high-speed vibratory range.

The greatest amount of data in both climbs anddescents was obtained at j.t= 0.15 and the Intrusion Indicesfor these cases are shown in Figure 8 as a function of theflight path angle. Flight path anglc (or shaft angle in awind tunnel test), has a relatively weak effect on

3.5

3.0

> 2.5"0

2.00

1.,5(tw

1.0

0.5

@ Level Flight ,,.o,.- Cw/


14/24

observedherehevibrationsthelowestwhilethelowestnoisesrecordedhereibrationssignificantlyincreased.tlowspeedhephenomenanvolvednBVInoisearedifferenthanhebasicphenomenanvolvednvibration.henoiseslargelyausedyparallelinteractionsfseparateipvorticeshedrompreviousblades,whilethevibrationsaresultoftheintertwiningof thetip vorticesttheedgeof therotordisk.Thus,dynamicistseedofocusheirexaminationsflow-speedvibrationotondescendinglight,butonthespeedregimeorpeakransitionibration.Two Unsolved Problems

A number of significant aerodynamic load problemsremain unsolved for articulated rotors and an advance isrequired in this area if any progress is to be made in theaccuracy of our predictions for loads and vibration. Theserepresent only single parts of a complexpuzzle-nonetheless, they are a good place to start.Negative Lift in High-Speed Flight

The airloads on the outer portion of the blade in high-speed flight are characterized by reduced lift at the end ofthe first quadrant and the beginning of the second quadrant,as is shown for the UH-60A and the research Puma inFigure 3. This negative loading starts to appear at Ix =0.25 or 0.30 and becomes progressively stronger asadvance ratio increases. It appears to occur on all rotorsfor which measurements are available (Ref. 22, 23).Detailed comparisons of calculation methods with theresearch Puma data (Ref. 8, 24) show relatively goodagreement with the measurements for this rotor.However, good agreement is not obtained for theUH-60A. Figure 9 shows a comparison from Ref. 25 ofthe measured section lift at 0.775R and 0.965R with thepredictions of 2GCHAS and CAMRAD/JA. Althoughthe amplitude of the negative lift appears approximatelycorrect for both analyses, each shows a significant phaseerror. Lim (Ref. 26) has examined this problem byexperimenting with options available in 2GCHAS, buthas not found modeling changes that resolve the phasedifference, although a number of the options changed theamplitude.The UH-60A model-scale data obtained in the DNW

tunnel (Ref. 27) show the same phase as the flight data(Ref. 28) and Sikorsky has examined this problem byreplacing portions of their analytical code withmeasurements (Ref. 29, 30). Using the measured airloads(Ref. 29), Torok and Goodman were able to show verygood results for the phase of flap bending moments,suggesting that the calculation of the aerodynamic loading

all,.11"m

..I

oo_)i:_1o4


15/24

theUH_OAhas twice that. Is twist, then, the importantparameter?Underpredietion of Blade Pitching Moments

A second problem that is a significant barrier toprogress in the prediction of loads and vibration is theinability of lifting-line analyses to predict the sectionpitching moments and hence the resulting control systemloads. Figure 10 shows the oscillatory pitch-link loads asa function of advance ratio for the research Puma (Ref. 8)and the UH-60A (Ref. 31). All of the analyses shownconsistently underpredict the oscillatory amplitudes overthe entire speed range of airspeed. The detailed timehistories for these calculations at high speed are shown inFigure 11. For the research Puma, the CAMRAD/JA andWestland/DERA analyses show reasonable agreement inthe qualitative form of the time history, but underpredictthe amplitude by about a factor of two. The originalCAMRAD analysis, however, does not show the correctqualitative behavior. For the UH-60A the CAMRAD/JAanalysis also fails to show the correct qualitative behavior.The inability of these analyses to predict the correcttorsional loading means that the torsional deformationwill also be incorrect and this will cause errors in theaerodynamic lift. There is no more fundamental problemin rotorcraft aeroelasticity than the deficiency in predictionillustrated here.

Maier and Bousman (Ref. 31) have examined thetorsion loading problem in considerable detail, takingadvantage of a number of experiments flown with theresearch Puma (Ref. 32, 33). Figure 12 compares theprediction and measurement of the two components oftorsional loading for this swept-tip rotor at _g = 0 deg.The section moment is just the aerodynamic momentabout the local quarter chord, that is, the moment cause bylocation of the airfoil center of pressure. The flightmeasurements show that this moment in increasing as theblade tip is approached and CAMRAD/JA does not predictthis effect. The lift-offset moment is a consequence of thesection quarter chord being offset from the pitch axis ofthe blade. For the research Puma the quarter chord firstsweeps forward and then aft. The CAMRAD/JA analysisshows a very good prediction of the section lift and hencethe lift-offset moment. It appears that the predictionproblem is largely one of calculating the unsteadyaerodynamic moments near the blade tip.

Lifting-line analyses become less accurate close to theblade tip and it would appear that the torsion loadingproblem would be amenable to CFD techniques. Bauchauand Ahmad (Ref. 34) have coupled the Navier-StokesOVERFLOW code to the DYMORE finite element modeland examined a moderate speed case for the UH--60A. Theresults for lift are unsatisfactory and it is not possible to

Flight TestCAMRAD/JACAMRADWestland/DERA

{q'Om,=4

+.L)

6.0

4.0

2.0

0.0

Puma (Fit 525)

, I , I , I , | , I0.0 0.I 0.2 0.3 0.4 0.5

_z

'o UH-60A (Fit 85)Z" 6 V

4.0

_'_ 2.0

0.00.0 0.1 0.2 0.3 0.4 0.5

Figure 10. - Measured oscillatory pitch-link loadscompared withanalyses for the research Puma and theUH-6OA; Cw/G = 0.08, 1-12 harmonics.

judge the moment predictions as an error was made in thecomputation (Jasim Ahmad, pers. comm.).

Not all practitioners have failed at this problem,however. Pavlenko and Barinov (Ref. 35) havedocumented the design experience of the Mil bureau forfour articulated rotors: the Mi-34 (3,400 lbs), the Mi-28(27,900 lbs), the Mi-8 (30,400 lbs), and the Mi-26(142,000 lbs). Figure 13 is a composite of variousfigures from Ref. 35 and shows the oscillatory torsion

11


16/24

4.0

2.0

0.0L)

-2.0

Flight TestCAMRAD/JACAMRADWestland/DERA

Puma,/2 = 0.362

I I I I I I I I45 90 135 100 225 270 315 360BLADE AZIMUTH, deg

UH-60A, /2 = 0.355

4.0

2.0

1. 0.0rj

-2.0

-4.0

-6.0

iV:# I I I I I I I I I

0 45 90 135 180 225 270 315 380BLADE AZIMUTH, deg

Figure 11. - Measured oscillatory pitch-link loadscompared with analyses for the research Puma and theUH-60A at high speed; Cw/G = 0.08, 1-12 harmonics.

Z.._..

-,


17/24

]0

O Oa)

F |

iI, i

IA,S, I_'n_

AMN.rmM

mU'

@'b

!..... I'

,::. ..-_c., \,-.'- I\w

/r_l, ""c) i,e_,

Figure 13. - Comparison of measurement and analysis ofoscillatory torsional moment for three Russian aircraft:Mi-34, Mi-28 (rectangular tips), and Mi-28 (swept tips)(Ref. 35) .

last 30 years in the accuracy of our prediction methods. IfI would identify a single, most important reason for thislack of progress, then I would say it is simply thetremendous difficulty of the technical problems that we aredealing with. However, nearly as important, in myopinion, are some cultural barriers that we have allowed todevelop. These barriers are so intimately tied into ourculture, how we work, how we find resources for ourwork, and how we communicate that work that it isoptimistic in the extreme to suggest that we can overcomethese barriers, largely self-imposed, in the near future.Nonetheless, it is worth a try.

The cultural barriers I refer to are varied and closelyinterrelated. I will focus on three of these barriers that Ibelieve have had the most deleterious effect on ourprogress and try to indicate changes that are needed. I willrefer to the three barriers as (1) the need for humility, (2)the need to use quantitative evaluation metrics, and (3) theneed to return to the scientific method. Each of thesetopics is so closely related to our own personalities andvalue systems as scientists that simply the process oflisting the topics can be seen as an insult to thecommunity. Nonetheless, arrogance intact, I will presson.

The Need for HumilityThe dynamics community has dealt with the

intractable problems of loads and vibration since the firsthelicopters and it does not appear that these problems havelessened in recent years. When one compares themultitude of papers that have been written in the last 30years that include the words "for the first time" and theabsence of papers that demonstrate improved prediction ofloads and vibration for helicopters, one has to ask whetherwe have even the most basic understanding of what theobjective of our research should be? Nothing can be moreexciting for a scientist then to tackle a difficult problem,break it into manageable pieces, solve those pieces, andput it back together. Less exciting, but importantnonetheless, is to work with one of those manageablepieces, and make real progress and share that with ourtechnical peers. But we need to remember the purpose ofwhat we are doing and that is to complete the synthesis,to bring the parts back together and demonstrate that theywork. And that is where humility becomes important.We need to understand that our part is only one piece ofthe puzzle. We need to understand that those that havegone before have given us the framework from which wecan build. We need to understand that the value of ourwork exists only in that it will be used and contribute tothe whole.

When I suggest that our community lacks sufficienthumility, I am not objecting to the personal demeanor of

13


18/24

/1Iq.ca

JdIG

ii"WO

-]_

@

-- . i 1"

1,',,ki l,ri i%1_ l , _V r

.i iill IO 1315 IN 21_

,, 'ilir"M

H.m I

a) ,AZ-IMU'3'rl, _ b)o 45

! :; i

I I"'",.,\ y1

Ill 140 li_ _ Jl.S J4l)_ZiMuT 4,rim;

x

_J

,J_ll |

Figure 14. - Comparison of measurement and analysis of torsional moment for three Russian aircraft at high speed:Mi.-34, Mi-28 (rectangular tips), Mi-8, and Mi-26 (Fief. 35).

researchers, but rather the words that they write. Iunderstand that one cannot perform research without funds,and one cannot obtain funds without a properrepresentation of why the work is important. But whenwe turn this into a competition to see who can hype andexaggerate the importance of their research the most--anddo this in our written papers, then we are harming ourcommunity. Let me give an example of what I consider ahumble approach. Regrettably, I' ve taken this examplefrom the CFD community; not from our own. In Figure15 I show the flight test data, the prediction of acomprehensive analysis, CAMRAD/JA, and a coupledprediction between CAMRAD/JA and the Full-PotentialRotor (FPR) code. A number of features from this figure(and the paper) are important. First, although FPR cancalculate pressure distributions on the blade and these weremeasured in flight test, this paper is not about pressuredistributions but is about aerodynamic loading--features

that are of primary interest to the rotor designer. Second,the figure compares the new results with data--it isimportant never to get too far from data. Third, thecomparison compares the prediction of the newmethodology with the current state-of-the-art, that isCAMRAD/JA in this case. The degree to which the newcalculations advance the state-of-the-art is an issue that isclearly illuminated in this paper. Not all of us can easilycompare our research with full vehicle measurements, norcan we always compare to the current technology. Ourfirst objective, however, should always be to compare ournew technology with measurement and comprehensiveanalysis and clearly illuminate what is the advantage wehave brought to the state-of-the-art. And when it is notpossible to make these comparisons, we need to qualifyour conclusions and point out the work that remainsbefore there can be a judgment of our new approach.

14


19/24

iFigure 15. - Comparison of computational methods withmeasured airloads on research Puma (Ref. 36); I1= 0.38.

The Need for Quantitative Evaluation MetricsTwenty-five years ago, at the Dynamic Specialists

Meeting at Ames Research Center, Dick Bennett said "...correlation, like beauty, is in the eye of the beholder."(Ref. 37) In my opinion this statement has become acrutch for the dynamics community. Now, anyconclusion about results, whether outrageous or simplyvague, is considered valid because we know thatcorrelation is not objective. Yet this statement has beentaken out of context. Dick said these words as an exampleof fuzzy thinking and in the next sentence he stated "Sowe must come up with a good definition, a workabledefinition of correlation."

Perhaps we have progressed so little in the last 30years because we are still waiting for someone to come upwith a good definition of correlation.., and are contentto continue our subjectively-based conclusions until thatday comes. I believe that we have failed to usequantitative metrics by choice, not because there is anydifficulty in establishing these metrics.

As an example, in calculating hover performance,Kocurek et al (Ref. 38) compared their analysis with flighttest data from 17 aircraft and defined an error band whichprovides a quantitative basis for judging the adequacy oftheir methodology. This approach was both systematic,in the use of so many data sets, as well as quantitative.Using a quantitative metric for a scalar value, such ashover gross weight may be simpler than establishing ametric for more complicated cases, but it wasaccomplished here because it was deemed important.

A first place to start for any quantitative metric is toplot the analytical results, in some form, as a function ofthe theoretical predictions. Good correlation is wheneverything lines up on a 45 deg line. I have illustratedharmonic correlation previously in this paper which issimply one method of applying this approach to theproblem. But it is only one approach in a vast field ofpossibilities. The important point is not to select one,agreed-upon, quantitative definition of correlation for theentire community but, instead, always include quantitativemetrics as a part of our research plans. Thus, before theresearch starts we need to identify the quantitative teststhat will be made to assess how good or bad the resultsare.

The Need to Return to the Scientific MethodEvery Philosophy Department in every university has

their Social Constructionists and these folks occasionallytalk eloquently about science and engineering. Theyaccept no basis for either, but view them as socialconstructions of society. Thus, the scientific method isnothing more than a cultural artifact put together by a partof society for its own benefit. As for engineering, it isjust a cut-and-try effort, accidental by and large, and againa vast set of rules or social constructs have been developedfor the use of the practitioners. The first time one reads orlistens to the exposition of a Social Constructionist onehas the tendency to dismiss them as jargon-driven fools.But it his hard to avoid some of their conclusions,particularly with regard to such icons as the scientificmethod. And it is particularly hard to refute their claimsif one is grounded in the literature of the dynamicscommunity.

What is wrong, in dealing with a very difficultproblem to propose a hypothesis, devise an experimentthat will test the hypothesis, and then either confirm ordeny the hypothesis? What is wrong with saying at thestart, that a quantitative method will be used to confirm orrefute the hypothesis? If there are no hypotheses to betested and no metrics, then everything we do is justphilosophy. Within the construct of the scientific methodan "experiment" might, in fact, be theoretical developmentrather than a physical experiment. But this does notchange the need to have the clear statement of thehypothesis and the means to test it. Wc admireresearchers who bring to our problems their brilliance andall encompassing imaginations. Yet if their research isdriven by ad hoc ideas that a new proposed functionality ora new scope for the analysis will at last provide the key tounlock the secrets of accurate prediction for loads andvibration, then I think we will face another thirty yearsabsent of progress.

15


20/24

Is There a Way Forward?I am hopeful that progress can be made in the next

three decades and that this community will not have tolisten to another frustrated researcher after another 30 yearshave past. Perhaps, real progress is near at hand, but mydim eyes cannot perceive this proximity. I think there arereal criteria that we can use to help us and many of thesecriteria are used by practitioners today. In order ofimportance my list of criteria are:

1) Research tasks should be selected where it is clearhow improvements, if proven, can be integrated in asynthesis step into one or more of our comprehensiveanalyses. Work on a reduced problem, if it cannot flowinto the synthesis step, is of substantially less value.

2) Research tasks should be selected where there is anobvious reference or benchmark, preferably based onexperimental measurements. Work should be discouragedfor which there is no means of testing the improvedanalysis or knowledge.

3) The research proposal should always be clear as tohow the improved analysis will be tested, that is, whatquantitative metrics will be used to demonstrate success(or failure). Multiple metrics are suitable; the absence ofmetrics is not suitable.

4) Whether a researcher incorporates the result of theirown research into the larger synthesis or leaves this taskto others, it is essential that all necessary steps be detailedso that a competent scientist can duplicate themethodology.

5) Where a research program develops experimentaldata, either for the reduced program or the full synthesis,then these data must be made freely available to theresearch community.

6) All of the inputs and supporting data used foranalysis should be freely shared.

There are a number of ramifications in my list ofcriteria and it is worth discussing these in detail. First,the openness that is inherent in some of these criteria,particularly the last three, eliminates participation in thislarger process by the industrial scientist. Proprietaryconsiderations, whether in respect to analytical orexperimental results, are of great importance to thetechnical health and future prosperity of each of ourindustrial concerns. The dynamics community cannotexpect industrial concerns to jeopardize their futurepotential for success. However, the industrial scientistdoes have an important contribution to make in a numberof areas. For example, it is one thing to developquantitative metrics, but in the end the specific numbers

that are selected must be relevant to the industrial process.Within industry there is a great deal of design experiencethat is valuable to all of us in trying to determine how toestablish quantitative metrics. We need to take advantageof this experience.

The last decade has seen the development of a numberof comprehensive analyses that are no longer internal tojust the companies. Thus, it is feasible to integrate newanalyses into CAMRAD II, CAMRAD/JA, 2GCHAS,UMARC (and its various versions), CHARM, andFLIGHTLAB. A barrier to use of these analyses is thedifficulty of establishing and checking out the input decks.Yet for many of the standard experimental problems agreat deal of validation of the input data has alreadyoccurred and it is pointless to repeat this work. Thus, thefree interchange of input decks is really a necessity if thereis to be reasonable progress in the future. Eventually, itmay not be too difficult to build input/output filters thatcan do most of the labor-intensive transformation betweeninputs for these various analyses. The more researchersthat use any of these analyses the easier it will be to focuson deficiencies in input data and correct them. Research isalways a competitive endeavor, but we should not becompeting at these elementary levels.

Progress will probably occur only if the communityis comparing multiple analyses over multiple data sets.Thus, to the extent data sets 'are restricted, the progress ofthe entire community is held back. In the experimentalarea, progress has meant a growth in the capability ofaccurate measurement as well as quantity ofmeasurements. Thirty plus years ago, pressure data fromten conditions seemed sufficient from a test of the CH-34rotor. All of these data were published in two slendervolumes containing tables of harmonics or time histories(Ref. 39, 40). Nowadays, the increase in the number ofparameters measured and the bandwidth that is used worksagainst any publication of the data and we are faced withenormous problems of how to explore, understand, andshare the data even in the absence of restrictions.

Just as the development and use of comprehensiveanalyses in the last decade has opened up newopportunities for the research community, it is apparentthat new experimental data sets that are coming along willalso provide a major opportunity to advance technology.Full-scale tests scheduled for the 40- By 80-Foot WindTunnel in the next five years may include the UH-60A,the AH-64A, and the V-22. Although only the first rotorwill include pressure instrumentation, the presence of adynamic balance on the Large Rotor Test Apparatus(LRTA) will provide unique vibratory load data that can beused to support research on vibratory loads. Anexpectation that useful data will be forthcoming to a wider

16


21/24

communitysprobablynrealistic.UnlikenthepastwhenhepublicationfdatawasamainstayftheGovernmentesearchacilities,imeshavechangedndmorehanikelythedatawillbe"owned"bythestakeholdershohaveobtainedt. Yetherealsosanopportunityorthewideresearchommunity.Increasinglyophisticatedeasurementsavebecomeincreasinglyoredifficultomakeandherestheopportunityorindividualesearchersrconsortiaodemonstratehatheycanenhancehevalueofthesedifficultexperiments.t thesameimeheymustnegotiateggressivelyoncerningheirstakentheresultingata.

Whatsneededntheutureoadvancehetechnologyfloadsandvibrationsastrangeeastndeed.Onewill needobemoreopenoopportunitiesonetworkwithcolleagueshethernindustry,heGovernment,racademia.newill alsohaveohaveagooddosefparanoiaokeepone'sfundingntact.OnemustbeatrueschizophrenicothatonecansellMach2,anti-gravityootsothepeoplewiththemoneynoneday,andhenextdaycomeoaforumsuchasthisandellthetruth.Finallyonemustdogoodesearch.h,if itwereonlythiseasy.References1. Piziali, R. A. and Du Waldt, F. A., "A Method for

Computing Rotary Wing Airload Distribution inForward Flight," TCREC TR 62-44, November1962.

2. Landgrebe, Anton J., "An Analytical Method forPredicting Rotor Wake Geometry," Journal ofthe American Helicopter Society, Vol. 14, No.4, October 1969, pp. 20-32.

3. Loewy, Robert G., "Summary Analysis," AGARDConference Proceedings No. 122, March 1973.

4. Piziali, R. A., "Rotor Aeroelastic Simulation--AReview," AGARD Conference Proceedings No.122, March 1973.

5. Sopher, Robert and Duh, James E., "Prediction ofControl System Loads in Level Flight andManeuvers," American Helicopter Society 50thAnnual Forum Proceedings, Washington, D.C.,May 11-13, 1994, pp. 1303-1321.

6. Hansford, Robert E. and Vorvald, John, "DynamicsWorkshop on Rotor Vibratory Loads Prediction,"Journal of the American Helicopter Society,Vol. 43, No. 1, January 1998, pp. 76--87.

7. Johnson, Wayne, "Recent Developments in theDynamics of Advanced Rotor Systems," NASATM 86669, 1985.

8. Bousman, William G., Young, Colin, Toulmay,Francois, Gilbert, Neil E., Strawn, Roger C.,Miller, Judith V., Maier, Thomas H., Costes,Michel, and Beaumier, Philippe, "AComparisons of Lifting-Line and CFD Methodswith Flight Test Data from a Research PumaHelicopter," NASA TM 110421, October 1996.

9. Bousman, William G. and Maier, Thomas H., "AnInvestigation of Helicopter Rotor Blade FlapVibratory Loads," American Helicopter Society48th Annual Forum Proceedings, Washington,D.C., June 3-5, 1992, pp. 977-999.

10. Kufeid, Robert M., Balough, Dwight L., Cross,Jeffrcy L., Studebaker, Karen F., Jennison,Christopher D. and Bousman, William G.,"Flight Testing the UH-60A Airloads Aircraft,"American Helicopter Society 50th Annual ForumProceedings, Washington, D.C., May 1 l- 13,1994, pp. 557-578.

1I . Johnson, Wayne, "Rotorcrafl AeromechanicsApplication of a Comprehensive Analysis," AHSInternational Meeting on Advanced RotorcraftTechnology and Disaster Relief Proceedings,April 1998, pp. 211-224.

12. Tung, Chee, Bousman, William G., and Low, Scott,"A Comparison of Airload Data Between Model-Scale Rotor and Full-Scale Flight Test,"American Helicopter Society 2nd InternationalAeromechanics Specialists' Meeting, Bridgeport,CT, October 11-13, 1995.

13. Wang, James M., "Challenges in Rotor DynamicsCorrelation," AHS International 54th AnnualForum, Washington, D.C., May 20-22, 1998

14. Brooks, T. F., Booth, E. R., Jr., Jolly, J. R., Yeager,W. T., and Wilbur, M. L., "Reduction of Blade-Vortex Interaction Noise Through HigherHarmonic Pitch Control," Journal of theAmerican Helicopter Society, Vol. 35, (1),January 1990.

15. Splettstoesser, W. R., Schultz, K.-J., Kube, R.,Brooks, T. F., Booth, E. R., Jr., Niesl, G. andStreby, O., "A Higher Harmonic Control Test inthe DNW to Reduce Impulsive BVI Noise,"Journal of the American Helicopter Society,Vol. 39, (4), October 1994.

17


22/24

16.Jacklin,Stephen.,Nguyen,KhanhQ.,Blaas,Achim,andRichter,eter,Full-ScaleWindTunnelestofaHelicopterndividualladeControlSystem,"mericanelicopterociety50thAnnualForumProceedings,ashington,D.C.,May11-13,1994.17.Swanson,tephen.,Jacklin,Stephen.,Blaas,Achim,Kube,Roland,ndNiesl,Georg,"IndividualladeControlEffectsnBlade-VortexInteractionoise,"Americanelicopterociety50thAnnualForumProceedings,ashington,D.C.,May11-13,1994,pp.81-102.18.Crews,SamT.,"RotorcraftibrationCriteriaANewPerspective,"mericanelicopterociety3rdAnnualForumProceedings,t.Louis,MO,May18-20,1987,pp.991-998.19.Anonymous,RequirementsorRotorcraftibration,

Specifications,odelingndTesting,"AeronauticalesignStandardDS-27,USArmy Aviation Systems Command, St. Louis,MO, 1988.

20. Mueller, Arnold W., Conner, David A., Rutledge,Charles K., and Wilson, Mark R., "Full ScaleFlight Acoustic Results for the UH-60A AirloadsAircraft," American Helicopter Society VerticalLift Aircraft Design Conference, San Francisco,CA, January 18-20, 1995.

21. Burley, Casey L., Marcolini, Michael A., Brooks,Thomas F., Brand, Albert G., and Conner, DavidA., "TiltRotor Aeroacoustic Code (TRAC)Predictions and Comparison withMeasurements," American Helicopter Society52nd Annual Forum, Washington, D. C., June4-6, 1996.

22. Hooper, W. E., "The Vibratory Airloading ofHelicopter Rotors," Vertica, Vol. 8, 1984.

23. Bousman, William G., "The Response of HelicopterRotors to Vibratory Airloads," JournalAmerican Helicopter Society, Vol. 35, (4),October 1990, pp. 53-62.

24. Bousman, William G., Young, Colin, Gilbert, Neil,Toulmay, Francois, Johnson, Wayne, and Riley,M. J., "Correlation of Puma Airloads-Lifting-line and Wake Calculation," Fifteenth EuropeanRotorcraft Forum, Amsterdam, The Netherlands,September 12-15, 1989.

25. Lim, Joon W. and Anastassiades, Tassos, "Correlationof 2GCHAS Analysis with Experimental Data,"

Journal of the American Helicopter Society,Vol. 40, (4), October 1995, pp. 18-33.

26. Lim, Joon W., "Analytical Investigation of UH-60AFlight Blade Airloads and Loads Data," AmericanHelicopter Society 51 st Annual ForumProceedings, Ft. Worth, TX, May 9-11, 1995,pp. 1156-1175.

27. Lorber, Peter F., "Aerodynamic Results of Pressure-Instrumented Model Rotor Test at the DNW,"Journal of the American Helicopter Society,Vol. 36, (4), October 1991, pp. 66-76.

28. Bousman, William G., "A Note on TorsionalDynamic Scaling," Journal of the AmericanHelicopter Society, Vol. 43, (2), April 1998,pp. 172-175.

29. Torok, Michael S. and Goodman, Robert K.,"Analysis of Rotor Blade Dynamics UsingExperimental UH-60A Airloads Obtained at theDNW," Journal of the American HelicopterSociety, Vol. 39, (1), January 1994, pp. 63-69.

30. Torok, Michael S. and Berezin, Charles R.,"Aerodynamic and Wake MethodologyEvaluation Using Model UH-60A ExperimentalData," Journal of the American HelicopterSociety, Vol. 39, (2), April 1994, pp. 21-29.

31. Maier, Thomas H. and Bousman, William G., "AnExamination of the Aerodynamic Moment onRotor Blade Tips Using Flight Test Data andAnalysis," Paper No. 48, Eighteenth EuropeanRotorcraft Forum, Avignon, France, September15-18, 1992.

32. Riley, M. J. and Miller, Judith V., "PressureDistribution on a Helicopter Swept Tip fromFlight Tests and from Calculations," Paper No.9, Ninth European Rotorcraft Forum, Stresa,Italy, September 13-15, 1983.

33. Riley, M. J., "Measurements of the Performance of aHelicopter Swept Tip Rotor in Flight," PaperNo. 35, Twelfth European Rotorcraft Forum,Garmisch-Partenkirchen, Federal Republic ofGermany, September 22-25, 1986.

34. Bauchau, O. A. and Ahmad, J. U., "Advanced CFDand CSD Methods for MultidisciplinaryApplications in Rotorcraft Problems," AIAA-96-4151-CP, 1996.

35. Pavlenko, N. S. and Barinov, A. Y., "Analysis ofTorsional Moments Produced in Main RotorBlades and Results Obtained," Paper No. V I.6,Twenty First European Rotorcraft Forum, Saint

18


23/24

Petersburg,ussia,ugust30-September,1995.36.Strawn,RogerC.andBridgeman,ohnO.,"AnImprovedhree-DimensionalerodynamicsModelorHelicopterirloadsrediction,"aperAIAA-91-0767,9thAerospaceciencesMeeting,Reno,NV,January-10,1991.37.Bennett,.L.,"Panel1 PredictionfRotorandControlSystemoads,"otorcraft Dynamics,

NASA SP-352, 1974, p. 306.38. Kocurek, J. David, Berkowitz, Lenard F., and Harris,

Franklin D., "Hover Performance Methodologyat Bell Helicopter Textron," Preprint No. 80-3,American Helicopter Society 36th AnnualForum, Washington, D.C., May 13-14, 1980.

39. Rabbott, J. P., Jr., Lizak, A. A., and Paglino, V. M.,"A Presentation of Measured and Calculated Full-Scale Rotor Blade Aerodynamic and StructuralLoads," USAAVLABS TR 66-31, 1966.

40. Rabbott, J. P., Jr., Lizak, A. A., and Paglino, V. M.,"Tabulated Sikorsky CH-34 Blade SurfacePressures Measured at the NASA/Ames FullScale Wind Tunnel," SER-58399, January 1966.

19


24/24

Form ApprovedREPORT DOCUMENTATION PAGE oMBNo.OZO4-O1S8Public reporting burden for this collechon of information is eat=mated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources,gathering and maintaining the data needed, and completmg and reviewing the collection of information Send comments regarding this burden est imate or any o ther aspect o f thiscollection of information, including suggestions for reducing this burden, toWashington HeadQuarters Services, Directorate for information Operations and Reports, 1215 JeffersonDavis Highway, Suite 1204, Arlington, VA 22202-4302, and to the Office of Management and Budget . PapenNork Reduc tion Projec t (0704-0188 ). Washing ton, DC 205031. AGENCY USE ONLY (Leave blank) 2. REPORT DATEMarch 20004. TITLE AND SUBTITLE

Putting the Aero Back Into Aeroelasticity

REPORT TYPE AND DATES COVEREDTechnical Memorandum5. FUNDING NUMBERS

6 . AUTHOR(S)

William G. Bousman7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES)

Aeroflightdynamics Directorate, U.S. Army Aviation and MissileCommand, Ames Research Center, Moffett Field, CA 94035-1000

9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES)

National Aeronautics and Space AdministrationWashington, DC 20546-13001 and U.S. Army Aviation and MissileCommand, Redstone Arsenal, AL 35898-50208

505-59-87

8. PERFORMING ORGANIZATIONREPORT NUMBER

A-0OV0012

10. SPONSORING/MONITORINGAGENCY REPORT NUMBER

NASA/TM-2000-209589AFDD/'FR-00-A-005

11. SUPPLEMENTARY NOTESPoint of Contact: William G. Bousman, Ames Research Center, MS T12-B, Moffett Field, CA 94035-1000; (650) 604-3748Presented at Eighth ARO Workshop on Aeroelasticity of Rotorcraft Systems. University Park, PA, Oct. 18-20, 1999.

12a. DISTRIBUTION/AVAILABILITY STATEMENT 12b. DISTRIBUTION CODEUnclassified -- UnlimitedSubject Category 02 Distribution: StandardAvailability: NASA CASI (301) 621-0390

13. ABSTRACT (Maximum 200 words)The lack of progress in understanding the physics of rotorcraft loads and vibration over the last 30 years

is addressed in this paper. As befits this extraordinarily difficult problem, the reasons for the lack of progressare complicated and difficult to ascertain. It is proposed here that the difficulty lies within at least threeareas: 1) a loss of perspective as to what are the key factors in rotor loads and vibration, 2) the overlookingof serious unsolved problems in the field, and 3) cultural barriers that impede progress. Some criteria aresuggested for future research to provide a more concentrated focus on the problem.

14. SUBJECT TERMSRotorcraft, Aeroelasticity, Vibration, Rotor loads

17. SECURITY CLASSIFICATION 18. SECURITY CLASSIFICATION 19. SECURITY CLASSIFICATIONOF ABSTRACT

15. NUMBER OF PAGES24

16. PRICE CODEA0320. LIMITATION OF ABSTRAC

Documents

Aero in Aeroelasticity