Sei sulla pagina 1di 20

Rotorcraft Analytical Improvement Needed to Reduce Developmental Risk-The 1989 Alexander A.

Nikolsky Lecture

Charles C. C r a w f o n l , Jr. Head, Aviation Tech17ologyBranch Georgia Tech Research Institute Georgia Institute of Technology Atlanta, G o .

Charles C. Crssford earned a Bachelors Dcgree in Aeronautical Engineering in 1954 and a Masters Degree in Aeronautical Eneineerine in 1970, both from the Georgia Institute of Technology. He served as a flight engineer . . an active duty as an officer in the Air Force and later as a civilian, between 1955 and 1962, at thc Air Force Flight Test Center, supervising the test of all classes af aircrafl. Mr. Crawfard transferred to the U.S. Army Material Command Headquarters in December 1962 to serve as Chief of the Technical Management Division and later as Deputy Project Manager for the Army's Light Observation Helicopter Program during the competitive development of the OH-4.41516. In May 1966, he joined the Army Aviation Systems Command, establishing the Flight Standards and Qualification organization to develop the Army's capability to qualify its own aircraft. This organization has been responsible far the development of Aeronautical Design Standards and for the planning and managing of airworthiness qualification programs far all Army aircraft. Principal qualification efforts include: AH-I Series, CH-47 Series, Bell Model 214, and Black Hawk, Apache, and AHIP designs, including the cnpines which they utilize. The organization was redcsignated the Directorate for Engineering with both development and qualification responsibility far new engines and aircraft as well as engineering support of fielded aircraft. Mr. Crawfard was a charter member of the Senior Executive Service, established July 13, 1979, and served as the Army Avistion Systems Command's first Technical Director. He was also the Director of the AVSCOM Research. Develonment and Eneinecrine Center until 1988. The Aviation Research and Technology Activity (with individual dirccturales en-luratcd ,,itl! UAS,\ Anne\, I ang1r.y imtl I . c v i ~ Kr\r.arch (:cntcr* n#~<l A+iotialn the AnnlirdTccl~nol~oer itnd l)cvelepmcnt .\rli\it!; . .. llircrlnrntc ul Furl Et#sti\,Virxinia); tttc Amty Atinnirs R t \ u ~ r e h IIBP Arm) ,%\ittion 1:nyinrcring Flight Acti,il); the 1)irrct~or.tlrs Enginrpriug, ,\dvanrcd Sjatonli, and !.ire for (:vrlr\ SnNuarr I<nginrcring; ontl thc 'lc.4 dnd Evalualiun managmnrnt nllicr, rcpanrtcd 10 hlr. Crnriord as the RD&E Center DirectorHe is currently the Branch Head of the newly established Aviation Technology Branch at the Georgia Tech Research Institute. The branch's activities include engineering support of the U.S. Air Farce Special Operations Farces helicopters and rotary wing technology efforts in the area of structural integrity, preliminary design, and experimental aerodynamics. Mr. Crawford has been a member of the American Helicopter Society since 1960, serving as Director at Large four terms, SecretarylTreasurer two terms, President, Chairman of the Board, and twice Chairman of the Vertical Flight Foundation. He was awarded an American Helicopter Society Honorary Fellowship in May, 1980. Hc received the Secretary of Defense Medal of Honor for Meritorious Civilian Service in 1988.

..

h e n 1 received notice o f m y selection b y the Society to present t h e 9th Alexander Nikolsky Lecture, I w a s indced startled. T h i s honor is normally awarded t o those stalwarts o f t h e Society that h a v e m a d e substantial theoretical conhi-

Preface

Nikolsky Lecture presented at the 45th Annual Natiollal F~~~ of the American Helicopter Society, Boston, Mass., May 22-24, 1989.

butions to o u r beloved technology. For someone of m y hackg o u n d to be so honored is very sincerely appreciated. Although I h a v e been a n active m e m b e r of the American Helicopter Society for 27 years, I never had the privilege of knowing Professor Alexander Nikolsky. However, his reputation i s familiar to most of o u r membership. A f e w have had the privilege of being his students. I a m quite indebted to o n e o f these. He i s Dr. Robin Gray of the Georgia lnstitute of Technology. Years ago he served as m y advisor in graduate

C.C. CRAWFORD, JR.

JOURNAL OF THE AMERICAN HELICOPTER SOCIETY federal government. Studies have been performed and elaborate reports written at virtually every level, from special Presidential commissions through the General Accounting Office, to special groups within the DUD and its military services. Many important suggestions of how to improve the developmcnt process have been compiled. A large maioritv of the new ideas have - " . been managerial in nature, such as "design freedom," "specification tailoring," "concurrency," and "program stability1 financial risk." Of equal importance have been the studics undertaken by the legislative branch of the federal government with the same objective, also resulting in constructive "management type" suggestions. Our Iiu;npean ;&I S o ~ i c countcrpdrts h;~vc t dcnionstratcd s vihrdnt and su.;ta~ncdinterest in rotorcr;~ftrcscarch and ~lcvclopment. There was a time when we could say that thc foreign rotorcraft interests were mainly ticd to the momentum of the U.S. activity. That is no longer true. Our nation is facing a greater technical challenge in worldwide aviation markets than ever faced before. This paper will suggest that the greatest contribution for reducing the risk of the developmcnt process may be a "hack to basics" technology initiative. In my view, this is the most critical missing ingredient. Its implementation rests with all, but the most painless birth of such a new initiative can occur by a recognition and appreciation of our ~roblems coming from the legislative side of'tke government. clearly, the benkts to the nation, both military and civil, are too important to sweeD the issue under the rug: Thc cver i,1ouari11g ialrre o rirne and f the deplerion of available real esrare will boost the imponance of vertical lift aircraft to the top of the aeronautical technology heap for both military and civil applications. 1 will illustrate the nccd for a "hack to basics" rotorcraft tcchnology initiative by first reviewing the multi-discipl~ncd t c c h n ( ~ l (c~ i~ r n ~ l c x ioirotary ~ i n ~ i i r c r n l i , to 111cnt1o11 ~~ tici not thc cvcr incrcaving technical co~nplcx~t~csol ~ S ~ N I ~ I I I I cquili~iicnt package integration essential f o r ~ a complete wcaponss~stem. However, a sound air vehicle is thc first order of busincss. Second, I will review by reference the higher risk elements of the design tcchnology and will compare the scopc of the preliminary design process for rotorcraft to their fixed-wing, but runway dependent, counterparts. Third, the shortfalls in the process will be illustrated by a review of the major design changes that occurred during the development of two major U.S. Army Aviation systems: the Utility Tactical Transport Aircraft System (UTTAS) and the Advanccd Attack Helicoptcr (AAH). Progress made in improving analytical tcchniques since their development will be pointed out. The need for further substantial improvement in this process is the capstone of this lecture. Sincc no "short tern-quick fix" is either logical or practical, 1 advocatc that the best course of action to insure viable military air items is the continued use of the proven competitivc "fly-off" acquisition strategy. 1 also plead that the work of developing and validating improved rotorcraft analytical methods be revitalized. Interdisciplinary Complexities of Rotorcraft I believe that all Society mcmbers would agree that the technical complexities and interdisciplinary requirements for rotorcraft far exceed those of fixed wing aircraft. In this regard, 1 mean no disrespect to the tremendous challenges faced by thc developers of the hypersonic "Orient Express" transport. However, the aerodynamic, aeroelastic, anddynamic interfaces of rotorcraft design represent the utmost challenge to our specialist. The three principle areas that are the toughest to predict during the design phase (thus requiring significant optimization in flight tests) arc vibration, loads, and aerodynamic performance. Later in the oaoer.. I will also build a case for significant risk in defining realistic empty weight for rotorcraft. Many aspapers have been written dealine with the interdiscinlina~ pects of rotorcraft design; however, 1 believe the &lurk has

school and has since frequently couns2led me on rotor wing technology. Of equal importance have been his frequent recommendations of outstandine voune academic ~erformers who thereafter have played key roles within our government in the business of advancine rotarv wine technoloev. Manv of the , previous Nikolsky lecturers have reviewed Professor Nikolsky's exceptional career and provide humorous accounts of specific activities. I will forego this, and simply point out that I sincerely believe that Professor Nikolsky, the renowned lecturer himself, would grin at the idea of a practitioner rather than a theoretician presenting a lecture in his name. Toward this end, I am truly grateful.
- <

Introduction The goal of this 9th Alexander A. Nikolsky Honorary Lecturership is to bring focus to what I believe is the most significant problem faced by vertical lift technology today: the need to design vertical lift aircraft that will "fly off the drawing board" without significant changes before achieving thc status of a viable operational vehicle. The problem is the shortfall in the realism and maturity of the analytical tools available to the designer for such an air item. The potential of vertical lift vehicles is almost without bounds. They were required in order to land on and return from the lunar surface. Of interest to this Society, however, is the more common use of low disk loading rotorcraft. The public acceptance of rotorcraft in roles such as an emereencv evacuation - , has grown dra~~~alically. many victi111\ol autc~mohilc For accidents, shipwrecks. ennhuuakes. and fires in hieh-r~sc hu~ldings; rotorcraft have providid the life line to survival. In military missions, they are required to position troops to occupy and hold terrain, which is the fundamental purpose of armies. This application, which started years ago through simple medical e k u a t i o n and observation hission<, has now expanded to the role of being a complete weapons svstem, with a wide range of firepowermade thrbugh integrated target acquisition systems operating over the full range of their respective sensor technologies. The economics of developing, . . - producing, and operating rotorcraft have been and will continue to he under constant scrutiny. The obligation to reduce the acquisition and direct operating costs of the& air items rests primarily with the design team; and using communities that set requirements and bear the burden of bcingproponents. Thc matter of improving the cfficiency of the development process clearly spans a larger group of organizations. In addition to develooers and users. a concerted effort to improve this process must'be supported by academic institutions and both the legislative and executive branches of the federal government, including the National Aeronautics and Space Administration (NASA). It is truly a shared responsibility of these parties to nurture the fundamental research efforts that will serve the public interest in defense and comnicrcial applications of the resulting technology. As an example-within the past couple of years, NASA conducted a study of the nation's wind tunnel facilities. The study considered the status and collective capacity of these The facilities to support future aeronautics re~earch'~ro&ams. study had NASA, industry, and Department of Dcfense (DoD) input. The results were presented to Congress with an agenda for renovation of these vital national assets. Congress approved and appropriated funds for the Wind Tunnel Revitalization Program. Specifically, the Full-Scale Aerodynamic Research Facility (40 x 80 wind tunnel) at the Ames Research Center was designated to receive modifications to enable aeroacoustic rotorcraft research. Therein lies a lesson on how the technical community and government can work together to address the problem of improving our national aeronautics research capability. Constant review and analysis of the weapons development process has been perfumed by the executive branch of the

~~~~~~

..

JANUARY 1990 been best portrayed in a single ch& by Dr. David Peters of Georgia Tech. Figure I contains some minor rearrangements of his work. I will use it for a brief discussion to set the stage for the problems of rotorcraft design and optimization. The basic dynamic components of a helicopter and their relationship to each other are shown graphically in this figure. It displays thc complex interdisciplinary intcractions that the rotorcraft technical community must accommodate. It considers only a single lifting rotor, but would be more complicated for multi-rotored configurations and anti-torque rotor systems. The upper three blocks represent the main lifting rotor. The rotor is divided into three basic theoretical models: blade aerodynamics, blade structural dynamics, and external flow field. These three theoretical models are not only highly coupled to each other, but are also coupled with the other areas of the dynamic analysis, airframe dynamics, airframe aerodynamics, power plantldrive system dynamics; and, of course, the flight control system. The complex interaction between these disciplines makes dynamic analysis one of the most difficult rotorcraft problems. The forces that drive the vehicle result from the angle of attack on the rotor blade airfoil sections. This angle of attack is made uo of three comoonents. One comoonent is duc to control enterei through the swasiplatc, another is blade due to the hlade motion itself. and the third is due to the flow induced by the wake. The angle of attack due to these components is used by the blade airfoil theory (which can range anywherefromquasi-steady aerodynamics, through panel models, all the way up to solving the Navicr-Stokes equations) to calculate thecirculation about, and airloads on the blade. These loads produce forces on the blades, feeding back to the blade structural dynamic equations.

'1989 NIKOLSKY LECTURE

These equations describe the blade motion when subject to the applied airloads, (which can be a difficult problem in itself if composite blades are used). This blade motion in turn feeds hack to the blade airfoil theory by contributing to the angle of attack. This feedback loop represents the inherent aeroelastic stability problem that highly flexible airfoils produce. In addition to the airloads, the blade airfoil theory also calculates bound circulation. This sheds a vortex wake, thus producing an induced external flow field. In addition to the lifting airfoil theory, and the blade structural dynamic theory, an induced flow theory (ranging from simple momentum theory to a complex free wake analysis) is needed to convert the time history of circulation into induced flow. These computations are extremely complex. A complete tightly coupled solution may not be practical for engineering application. Thus, the coupling between those must be carefully adjusted to achieve an acceptable balance between accuracy and computational efficiency. In addition to high coupling among the three major rotor dynamic theories, the rotor dynamics are also coupled with fuselage dynamics. The blade structural dynamics is transmitted to the fuselaae in the form of hub loads-the shears and moments whichdrive the fuselage motions by way of the transmission mounts. A similar condition exists through the engine mount. The dynamic response of the airframe to these loads feeds back into the root of the hlade as boundary conditions. For this reason, it is not accurate to calculate loads for a fixed huh and apply them on a fuselage. The shaking of the fuselage from the rotor loads in turn shakes the rotor, and that changes those loads, resulting in a change to the root chord boundary condition. (I guess the cliche, "What goes around, comes

DESORED FLOGMU GOMDWOOMS


WAKE IMPINGEMENT

CIRCULATION

STRUCTURAL DRIVE TRAIN

TAIL ROTOR &OTHER CONTROLS

WAKE IMPINGEMENT QUASCSTEADY RIGID BODY MOTIONS

AERODYNAMICS &T H R U ~ ~ E R ~ SURFACE MOTIONS

FUSELAGE WAKE

Fig. 1. Rotarcran multi-discipline interactions (for main lifting rotor).

JANUARY 1990 tremely complex. Also, due to thz structural complexity of most airframes, a large and detailed finite element model is required. Programs l k e Dynamic Analysis Methods for Vibrations (DAMVIBES), sponsored by NASA's Langley Research Center and i n v d ~ v i n ~ principal helicopter fou; manufacturers, are helping to broaden our understanding of fundamental dynamic finite element modeling principles, and improved correlation between predictions and test results for furelaee natural freouencies has been obtained. However. the ~ ~ ~ ~ ~ cost and time requi;ements of dynamic model development is hieh. and this will onlv he compounded as composite materials -. make up more of the airframe's structure. The maior source of rotorcraft vibratorv excitation is, of course, the rotor system. The complication of rotational inertial forces. the flexihilitv of rotor blades, the complex aemdynamicsbroduced by h e rotor system and interfcience with ihe airframe, and coupling of the system with both airframe and powerplant dynamics-all these make rotor system dynamic analvsis tough. Charles Niebanck's candid paper (Ref. 3) on thc i;ttcmplei appli~,ation hlade inays tuni"b.t#, re(lucc rolo~ of vibratory loads demonvlraler the nccd for more accurale analysis capsbilities. In that paper, a blade with analytically optimized mass distribution (optimized to give the minimum normalized magnitude of the sum'of all six huh loads components) was compared with a production blade. A comparison of S-76 flight tests results between the tuned blade and the production hlade is given in Fig. 4, and shows that the tuned hlade actually produced (higher) vibratory loads-obviously better analytical tools are required. The reason that rotor loads accuracy is such an important part of vibratory can be seen in Fig. 5 (Ref. 2). The airframe response to rotor shaft loads receives a contribution from manv load comoonents. and an error in one component will result in an erroneous vihration prediction; the cumulative error in all the load components is usually too large to allow accurate vibration predictions. Yet another reason for the difficulties in vihration analysis is the need for accurate higher harmonic loads. The technical complexities of predicting rotor loads is discussed further in the next section. Although Loewy pointed out that great reductions in vihration levels have hcen made over the past 40 years, see Fig. 6 ,

1989 NIKOLSKY LECTURE


COUPLED ELASTIC F-E INERTIA

PHASE, deg

Fig. 5. Calculated 3lrev lateral shear for baseline blade; V = 160 knots. this has resulted from brute force vibration devices, rather than improved prediction methodology. Vibration analys~sis still more of an art than a science. Thus, virtually every new helicopter demonstrated higher than predicted vibration levels early in flight test programs. This led to extensive trial-and-error efforts during flight test to reduce vibrations to quasi-acceptable levels. It is most important to note that vibration differences in actual flight have been key factors in the evaluation of prototypes in every "competitive fly-off" by the military. Until a better grasp of the problem is reached, designers are forced to rely on basic factors such as proper placement of natural frequencies for the rotor blade, drive system, and fuselage, and adequate fuselagelrotor aerodynamic clearances. In the future, a combination of this plus extensive wind tunnel testing, and the use of new vihration control devices and concepts, followed by optimization of the whole system in flight test, will be needed if vibratoly levels are to be lowered to obtain a more reliable weapon system. Rotor Loads Predictions After vibration prediction, rotor loads prediction is the highest technical risk area of rotorcraft development. Indeed, the inabilitv to accuratelv oredict rotor loads is a maior reason for <. the difficulty in vibration analysis. The state-of-the-art rotor loads orediction methods received much attention in the 1974 (Ref. 6) hypothetical helicopter rotor analysis comparison, where several maior airframe manufacturers compared predictions of natural frequencies of a given rotor in a vacuum: It was found that, even for this relatively routine task, there was significant disa;reemcnt bctueen thc compan~cs.William ~ou\man'sreview of rescarch i n rotor loads (Kcf. 7) d < ~ c \ excellent jab an oicovcring the progress that has hecn madc qincc 1974, showing both vrcas of gaud agrucmcnt and arcds where correlation ispoor. It is concluded that even though advances have been made in rotor loads analysis, there is still a pressing need to develop our understanding of basic physics of the rotor system. Predicting rotor loads is a more tractable problem than predicting vihration levels in the fmelage for several reasons. First of all, the rotor loads are dominated by the steady component

4P A i r f r a m e A c c e l e r a t i o n 100%N,

INDICATED AIRSPEED. KNOTS


Fig. 4.

Tuned blade night test results.

C.C. CRAWFORD, IR.

JOURNAL OF THE AMERICAN HELICOPTER SOCIETY

0.7

0.4
CABIN VIBRATION LEVEL (g)

03 .

0 1950 1960 1970


YEAR
Fig. 6. Trend of helicopter vibration (relative to mil specs). and the first oneor two harmonics, whereas the higher harmonic loads are more important for vibrations. Most of the dynamics which are not closely related to the rotor system, such as the fuselage dynamics, powerplant/drive train dynamics, and automatic flight controls, are expressed in a reduced order model (if at all) for rotor loads analysis, and the state-of-the-art now is such that only rotor aerodynamics, rotordynamics, and sometimes the wake, are integrated. Also, the finite elemont modeling of the airframe does not play as critical a role as it does in the vibration analysis. Although analytical computer programs, such as Bell. Helicopter Textron's C8l (Refs. 8 & 9), have made great improvements over the last decade, the accuracy of even the most sophisticated programs available may not be good enough for the technical challenges the future rotocraft will demand. As pointed out by Dr. Dan Schrage in his Overview of Technical Problems in Helicopter Rotor Loads Predictions Methods (Ref. lo), maximum sustained cruise speed (the flight condition most often used by industry to substantiate fatigue design) shows relatively poor correlation with flight test. See Fig. 7. The problems in rotor loads analysis can arise from errors in the rotor system modeling (rotor aerodynamics, dynamic, and wake), errors in the integration of the rotor system with the rest of the rotorcraft (hub modeling, aerodynamic interferences), or even the mathematical solution method. The rotor aerodynamic and wake analysis face challenges such as dynamic stall, airframelrotor aerodynamic interaction, blade vortex interactions, and the pitching dynamics, to name a few. A lack of agreement on the basic aerodynamics of a rotating airfoil, such as three-dimensional stall, reflect the need for additional research in the fundamentals of rotor systems. ~ e blade tips and airfoil sections, and highly cambered blades, w will introduce problems with thecurrent analysis methodology. An example of such is the coupllng of the torsional modes with the flap bending modes which can result in blade dynamic twist ("Mach Tuck") as discussed by Schrage (Ref. 10). Such phenomena indicate. that extensions of rotor loads analysis methodology will be necessary to support new rotor systcm configurations. ~ a i u r to account properly for interaction of the rotor system e with other facets of the rotorcraft often introduces crrors in the rotor loads. In order to simplify the airframe interaction, the hub is either assumed fixed or is modeled with some simplified analysis. If a complete finite element model were to be used,

1980

1990

2000

50

100

150

200

250

300

350

BLADE STATION (IN)

Fig. 7. Comparison of AH-64 predicted and measured flatwise moment diitributian.

JANUARY 1990

the problem would become too computationally intense, but determination of correct blade root boundary condition model is not well understood. Yen and McLarty (Ref. I I) have shown the necessity of modeling the pylon impedance, as shown in the OH-58A loads in Fig. 8. Also, rotor configuration differences such as teetering, articulated, or bearingless, makc dynamic dissimilarity a problem. Thc dynamic behavior of thcse different configurations makes different aspects of the problem important for each. New configurations are hard to analyze due to composite materials (for which elastic coupling terms become important), branched load paths, or other aspects of novel design concepts. Finally, the solution method is also a potential cause for erroneous rotor load predictions. Problems with assumptions in the "negligible" coupling of bladc modes, the problems introduced by load discontinuity at the blade root, and the lack of understanding of how many blade modes are needed for an accurate solution all give the mathematical solution additional comolcxitv. The importance of accurate rotor loads prediction is seen in the strone imoact it has on rotor svstem weieht and fatigue life of dyna;ic c&nponcnts. If the loads are under-predicted, thc rotorcraft will havc either a restricted flight envelope or shortened retirement lives of dynamic components; if they are overpredicted, the rotorcraft will be too heavy to perform its mission

1989 NIKOLSKY LECTURE

efficiently. Errors in rotor load predictions result in the need for redesign of fhc rotor dynamic components, usually after first flight (see Design Changes Section.)

Aerodvnamic Performance
The next most difficult aspect of rotorcraft prcliminary design is the aerodynamic prediction. I use that term rather than flFght to separate the impact of empty weight prcdictions. Estimation of the powerplant installation losses is usually within reason. The aerodynamic prediction problems were addressed by Frank Harris in his 1987 historical pers~cctive nresented at a National Suecialists Meeting of Acrodynamic; and Aeroacoustics (Rei 12). This pap& superbly states the deficiencies in helicopter performanccestimationtheory and provides penetrating insights into the following operating regimes: hover and vertical rate of climb prediction; flapping and feathering trim and rotor thrust in forward flight; and pcrformancc with rceard to rotor ~ r o f i l e -. rotor induced drae. drag. and blade stall onset. Hover performance data as shown by Harris demonstrate that the non-dimensional performance of four hclicopters designcd for the Army in the mid-1970's were essentially the same when tail rotor was excluded and adjustments were made for solidity differences. The initial prediction and the actual flight measurcment were significantly diffcrcnt. These differenccs arc repeated here in Fig. 9 for completeness. The flight data were taken during ~ r m ; competitive tests (Refs. 13; 14, 15, 16) and measured as accurately as possible at the timc. For a more recent case involving a modern technology rotor (OH58D), the predicted non-dimensional hovcr performance was 9 pcrcent ootimistic (Ref. 171: and.. for the most recent new com.. posite;otor blade developed for the Army's use with the UH1H fleet. the resultine hover lift was 3.4 oercent low (Ref. 181. Add to this short?all in predicting dower and the weight correlation for the hover case, a further limitation in predicting the vertical rate of climb. However, thc power to climb is sisnificantly less than momentum theorv would lead the designer to bllieve. Figure 10 from Ref. l i illustrates this point. In a time frame in which vertical rate of climb has such tactical importance to the military, and low empty weight is of such political importance to support arguments for lower acquisition costs; the absence of adequate prediction methodology is unacceptable. This is aggravated by the fact that the flight test techniques for measuring vertical climb arc so difficult that its measurement is a treacherous situation. Notc the scatter in the test data of Fig. 10. For assessing the forward flight condition with trim conditions. the aerodvnamics communitv is unablc to oredict accurately blade flapping and rotor thGust. In assessini how good simole momentum theorv is. Harris made com~arisons theof orehcal predictions with-test data gathered by 'Mc~loud, Bigeers in Ref. 19. In this referenced reoon, the first-harmonic flapping was brought to zero by adjusiments of cyclic control positions. This enabled correlation with over 1,500 data points, over many advance ratios and Mach numbers for seven different rotors. When these tests are submitted to rigorous computer analysis for correlation of theoretical and test results, then; and only then, can the aerodynamics community demonstrate the adequacy of today's theoretical methods. By reference, Hanis also shows that frec wake analysis shows a better "theory-test correlation" for rotor thrust prediction in trim flight. Given today's computer code capabilities, Harris also advocates usine actual non-uniform induced velocitv to compute rotor thrust trim for forward flight trim conditions. To accuratelv comoute rotor oerformance for hieh advance ratios, focus is required on three interrelated factors, all of which limit rotor performance: (1) Rotor Profile Drag (2) Rotor Induced Drag (3) Retreating Blade Stall Onset

. .

---- HUB FIXED -BLADES WITH NOMINAL IMPEDANCE AT HUB

~~~

CHORD

Fig. 8. Comparison of OH-58A blade bending moments with and without modeling of pylon impedance; V = 83 knots.

10
YAU-M

C.C. CRAWFORD, JR. *

JOURNAL OF THE AMERICAN HELICOPTER SOCIETY

Fig. 9. UTTASIAAH hover OGE performance (predicted vs. actual).

Another theoretical problem area identified for the aerodynamics community is accounting for induced drag at advance ratios greater than 0.25. The areas where helicopter aerodynamic theory and actual behavior do not correlate have been eloquently described by Harris in much greater detail than I am comfortable in discussing. However, I must emphasize that the problem areas identified and discussed by Harris are consistent with my thcme of buildirra stronaer rheorelical methods to ensure that rotorcrafrperf&n aspredicted, without significant changes during the develo~mentorocess.. onlv simde outimization. We can . then be realistic in seeking greater advances in helicopter performance. M a j o r Air Vehicle Design Changes in Flight Test Before I review specific design changes, I think it appropriate that I first review the scope of each development phase in terms of number of flight assets and actual flight hours utilized in the developmentlqualification process. Both the Army's UlTAS and AAH programs consisted of a "competitive fly-off" phase emphasizing the capability of the basic air vehicle followed by a maturity phase which fully optimized and qualified the complete weapons system. In the case of U'ITAS, the overall development spanned seven yean and nine months (from award of the development contract to the approval to enter full rate production). The scheme of this development is shown in Fig. I I. The basic engineering development utilizes three prototype helicopters accumulating approximately 700 contractors' flight test hours and approximately 750 government engineering and operational test hours. The government test effort included that required for the "fly-off," as well as the all important conventional airworthiness assessments. This does not include a fourth company-owned prototype which supplemented the government funded development pro-

gram, hut was fabricated relatively late during the first phase. In the initial funding request to the Congress, the Army had asked for five prototype aircraft in addition to a Ground Test Vehicle (GTV). The Congress restricted the funded program to the three prototype articles mentioned plus the GTV, however, the companies found it necessary to construct a fourth flight article with their own funds to ensure adequatc dcvelopmental hardware. Although they came on the scene much later than would have been desirable, these company-owned prototypes played a key pmhlem-solving role. The GTV was utilized for over I200 hours of tie-down operation. During the maturity phase for the winning UH-60A, an additional 600 hours of contractor flight testing was accumulated on the prototypes, plus 150 flight hours on the first two production aircraft. Government testing during this phase included another 150 flight hours utilizing the prototypes and 1100 fl~glits utilizing early production aircraft. GTV operations continued for another 700 test hours. The totals: 1450 hours-contractor, 2000 hours-government, and 1900-hours GTV. In the case of the AAH, a "competitive fly-off" was also used, however, the scope of the developmental effort prior to the "fly-off" was significantly less. It relates to what today would he considered a Demonstration/Validation effort. The scheme of the AAH development effort is shown in Fig. 12. Phase 1, the developmental work prior to the Government Comprehensive Test (GCT), was primarily air-vehicle oriented. Weapons operation was limited to a rough cut assessment for structural implementation. Phase 1 utilized two prototypes accumulating approximately 300 flight test hours and approximately 100 govemment engineering flight test hours. The government operational test effort was only 25 flight hours. A GTV was utilized for approximately 365 hours of powered tiedown operation. During Phase 2, which constitutcd Full Scale Development (FSD) for the winning AH-64, three additional

JANUARY 1990

1989 NIKOLSKY LECTURE

II

MOMENTUM
'HOVER

FAIRING FOR OH-58D

Fig. 10. Momentum theory overprerlicts power required for vertical elinah.

prototype aircraft were manufactured. These were primarily for develovmental and intenration of mission eauivment which incorporstcd a new ~ a r g ; Acquisition and designation System (TADS)/Pilot Night Vision System (PNVS), a laser semi-active guided point targct anti-armor missile (HELLFIRE) and a new 30mm arcd wcapon. The contractor's Phase 2 flight test effort encompassed an additional 980 hours of testing utili7,ing the two original prototypes and 715 hours of tcsting using the three new prototypes. Thc total government flight testing, enginccring and operational, on all prototypes during Phasc 2 was 880 1 hours. This included 200 hours for Operational Test 1 (OT 11) which was the key prercquisite for a production decision. Continued GTV operation of over 1050 hours contributed significantly to the quality and reliability of the rotor-drivc system. The loss of one test aircraft due to a mid-air collision, and early Phase 2 budget instability, clearly impacted the overall schedule. The final airworthiness qualification, using early production aircraft, included 170 contractor flight test hours and 140 governmcnt engineering flight test hours. The period between the initial contract award and the decision to enter production encompassed eight years and six months. The flight totals are: 2175 hours-contractor, 1145 hours-government, with 1415 hours-GTV. As mosi will recall, the design changes required after first flight were significant. In some cases, changes needed to meet Army requiretncnts could not he incorporated in the relatively short time available to prepare for the fly-off. Therefore, some of the required changes were simply "proposed" for the production configuration with limitcd analyticallwind tunnel suhstantiation orovidcd as Dart of the nroduction orooosal. It is interecting to notc that those de\igns 11131 ~~~corpor:itetI improvemcnts hcforc the tlv-oft !rcrc 1011kcduoon ;I.; hav~ne s~ell~flc;tnt lower risk of ultimately meeting minkurn ~ r nee&. ~This m

. .

was particularly true in the areas of vibration, flight performance. and flving aualities which were all immediatclv ao. -. parent to Army evaluation pilots. The YT700-GE-700 e&ink, common to all designs, performed exceptionally well during the entire "fly-off" process due to its relative maturity. This had resulted from its exhaustive developmental tcsting prior to installation in an aircraft. Thc design changes to he discussed herein have been limitcd to those areas whcre the change may have been avoided if a more encompassing preliminary design phase, bettcr analytical methodology and substantially more wind tunnel testing and simulator activity, had been available. Adequate analytical tools were simply not available; nor did time permit the othcr activities. The UTTAS schedules were naively belicvcd to be realistic but the more stringent AAH development schedule fclt the urgent pressure of the need for a more capable attack aircrafi; following the cancellation of the AH-56, Cheyenne. While affordability was the most visible reason to thc public for thc AH-56 cancellation, technical problems exasperated by the lack of adequate analytical methodology were also a major contributor. For thc record, the most significant design changes together with the problems uncovered in flight tests that dictated the change are outlined in Figs. 13, 14, 15, and 16. It is interesting to note the types of changes that are common. These arc: (a) Increases in thc main rotor height above thc fuselageto minimize vibration and aefodynamic interference between the rotor and fuselage; and otherwise enhance safety through increasing the clearance over the canopy, the tail boom, and ground personnel. (Fig. 17 illustrates the dramatic differences for the UTTAS vehicles.) (h) Revised empcnnage configurations-a widc variety of horizontal tail appendices; some fixed, some movable, all

12

C.C. CRAWFORD, JR.

JOURNAL OF THE AMERICAN HELICOPTER SOCIETY

Fiscal Year

72

73

74

75

76

7T
A

77

78

79
A

80

8 1

A 1 DSARC I & I! MAY 71

A 1 DSARC Ill NOV 76

ASARC llla OCT 79

V FIRST FLT OCT 74


B SC ENGR OEVEL (BED) AI

I
MATURITY PHASE

AlRFHAME

PROOUCTION SOURCE SELECTION


FIRST FLT NOV 7

~
T F I R S T FLT OCT 78

BASIC ENGR OEVEL (BED)


YUH-6IA

I PEP PRODUCTION
FOTE PVT
PTU

m GOVT

A I OSARC . . ARMY I DEFENSE SYSTEM AC(1UISITION REVIEW COUNCIL GCT . . . . . . . GOVERNMENT COMPETITIVE TEST PEP . . . . . . . PROOUClBlLlTY ENGINEERING & PLANNING FDTE . . . . . . FORCE DEVEL TEST & EXPERIMENTATION PVT . . . . . . . PRODUCTION VALIDATION TESTS
Fig. 11. UTTAS program structure.

relocated; and revised vertical fin areas to enhance tail rotor thrust and reduce blockage. (c) Main rotor blade configuration changes in the form of either an airfoil section change, a blade planform change, or both-to improve high speed maneuverability and eliminate "Mach tuck." (This is a very large nose-down pitching

moment on the advancing blade in high speed flight which produces high loads and increased vibration.) Only the YUH61 did not rcquire such changes, in that another type of analytical breakdown occumd-the high speed lift capability of the rotor had been underestimated. (d) Tail rotor configurations-either physically changed,

Calendar Year
DSARC I SEP 72
'

73

74

75

76

77

78

79

80

8 1

82

83

84

85

89

ENGR DEVEL (COMPETITIVE)YAH.6$ PHASE II FULL SCALE ENGR DEVEL

CONTRACT AWAROYUN PHASE I YAH-63

II DEC t 6 A S A d C Ill N b V 8 l t VDSARC Ill MAR 82'

,
I

TCONT~T AWARD DEC 76 TADS SELECTION APR 8 0 1

JUN 81 OT II W AUG 81
V

I I

CONTRACT AWARD DEC 81VFIRSTFLT DEC


PRODUCTION

LONG LEAD FEB 81V

-[,I
AUG 84 OCT 84

FIRST ARTICLE
PRE.PRODUCTION TEST

SEP ~3~ TT YEB 85 i PVT CONTR G V OJ !

m B

A IDSARC . . ARMY IDEFENSE SYSTEM ACUUlSlTlON REVIEW COUNCIL GCT . . . . . . . GOVERNMENT COMPETITIVE TEST OT . . . . . . . . OPERATIONAL TESTING PVT . . . . . . . PRODUCTION VALlOATlON TESTS TAOS . . . . . . TARGET ACUUlSlTlON & DESIGNATION SYSTEM
Fig. 12. A A H program slructure.

JANUARY 1990

I989 NIKOLSKY LECTURE

13

their control authority changed, or their operating environment changed for the purposes of improving sideward flight capability to ease pilot workload when hovering in gusty wind conditions. (e) Numerous fairings were added and components realigned for the purpose of reducing external drag. A more cost effective developmental program would extcnsively address this subject during eady wind tunncl testing. 1 will say more about the exhaustive wind tunnel test performed on Boeing's commercial 757 and 767 later in the paper. Some optimization in tailoring of flight control systems will always be required during the flight test phase. Future approaches will minimize such flight testing through theextensive use of simulators durine thc ~reliminarvdesien ohase. How.. ., . ever, lhcse s~rllulat~oas he no inure accuratc than thc anwill alvlical tools which nroducc thc st:~bilitvderivatives :~nd control analogs elementary to their operation. Vibration reduction will always be a major thrust of rotary wing flight tcst; howcver, better analytical methods would make possible a more rapid understanding of the surprise sources and vibratory levels encountered in early flight tests. Such an understanding leads to quicker and more cost-effective corrections.

the lack of adequate wind tunnel testing during the preliminary design phase. Though more complex than their fixed wing counterparts, very little effort was spent on wind tunnel testing. The testing done both pre- and post-first flight is shown below. In some cases, more wind tunnel activities occurred after the aircraft was in flight tcst than were performed during the initial design phase.

'AH-63

YAH~64
514 432
-

PRE-FIRST FLIGHT Airloil Sections Airvehicle Full Scale TIR

Subtala1
POST-FIRST PLlGHl

84 293 50 425
-

946

Airroil Sectialls
Airvehicle Full Scale TIR
Sublolal

1615 97 1712
NIA 4045

1470 160

1630 455

96 321 24 45 1
NIA 876

886
-

886 623

PRODUCTION GRANDTOTAL

2715

1832

Wind Tunnel Testing


Not withstanding the lack of maturity of the analytical techniaues used in the desien of comnetitive rotorcraft for the Arm i ' s AAHIUTTAS p&grams p;eviously discussed, another reason for the large number of configuration changes that were required of each design after it entered flight test was due to

With the exception of full scale isolated tail rotor tests, the wind tunnel testing was accomplished utilizing small test mod-

TOPUG
BASIC AIRFOIL .MAIN ROTOR BLADE ROTOR HEIGHT eSTABILATOR CONFIG

FRT TEST PROBLEM


MAX LOAD FACTOR LOW OHOVER CAPABILITY LOW HIGH VIBRATION HlGH ROTOR BLADE LOADS *AIRCRAFT ATTITUDES IN LOW SPEED FLT REGIMES HIGH SHAFT BENDING LARGE TRIM CHANGE UNSTABLE SIDEWARD FLT HIGH FLT LOADS

DESUGN CHANGE
ADDED DROOP SNOOT TIP CAP EXTENDED 4 INSBLADE ROTOR MAST INCREASED 15 INS LOW FIXED STABILIZER RAISED MADE MOVABLE OVER A 43 DEG RANGE THRU AUTO CONTROLS REDUCED FIN AREA WITH CUT-OUT ROTOR SHAFT, SUPPORT STRUCT, & CONTROLS REDESIGNED FOR INCREASED STRENGTH CHANGED INCIDENCE ANGLE TO ALIGN WITH FUS FLOW REPLACED GREASE LUB WITH OIL LAT SENSITIVITY INCREASED YAW/ROLL MECHANISM INCORPOR REDESIGNED FOR INC STRENGTH MOVED INBOARD DAMPER OUTBOARD BY 3.5 INS INCREASED SIZE & PRESSURE TO INCREASE STABILITY MARGINS FLEW OVER 100 FLTS TO OPTIMIZE REDESIGNED TO IMPROVE AIRFLOW PATH CHANGED FROM FLUIDIC TO ELECTRICAL/MECHANICAL

VERTICAL FIN AREA .TAIL ROTOR MOUNT & CONTROL LANDING GEAR SPONSON TAIL ROTOR GEARBOX CONTROL SYSTEM AUTHORITY .MAIN ROTOR SERVOS MAIN ROTOR DAMPERS .MAIN TIRES

HIGH DRAG UNDETECTED OVERTEMP POOR CONTROL HARMONY HIGH FLT LOADS HIT DAMPER STOPS IN HlGH SPEED AUTOROTATION MECHANICAL GROUND INSTABILITY EXCESSIVE ERROR HIGH TAIL VIBRATION & BUFFET VARIABLE FLT CONTROL SYS CHARS W/ EXTREME TEMPS

AIRSPEED SYSTEM .AFT NACELLE CONFIG .STABILITY AUG SYS

Fig. 13. YUH-60A required design changes.

14

C.C. CRAWFORD, JR.

JOURNAL OF THE AMERICAN HELICOPTER SOCIETY

TOPUG
ROTOR AEROELASTIC STAB

FLT TEST PROBLEM


EVAL WITH/WITHOUT DAMPERS (INITIALLY REQUIRED BY ARMY) REDUCED SIDEWARD FLT MANEUVER CAPABILITY YAW "KICKS" IN SOME FLT COND RAN OUT OF CONTROL DURING LOW SPEED MANEUVERING TOO SLOW FOR MANEUVERING

DESIGN CHANGE
LAG DAMPERS REMOVED INCREASED DIA FROM PROPOSAL, FURTHER INCREASE FOR PROD ADDED BEANIE LARGER FAIRING FOR PROD DOUBLED LAT SWASHPLATE TRAVEL DOUBLED CONTROL ACTUATOR MAXIMUM RATE REDUCED FIN CHORD ADDITIONAL FUS ABSORBERS EXTENSIVE TESTING OF ABSORBER INSTALLATIONS MANY LOCATIONS TESTED FLEW FIXED HALF SlZE TAlL PROPOSED SAME FOR PROD

TAIL ROTOR SIZE

ROTOR WAKE INTERFERENCE ON TAIL ROTOR LATERAL CONTROL TRAVEL

FLT CONTROLS RATE LIMITED

VERTICAL FIN SIZE

EXCESSIVE T/R BLOCKAGE IN SIDEWARD FLT EXCESSIVE VIBRATION

VIBRATION ABSORBERS

PlTOT STATIC SYSTEM HORIZONTAL TAIL SIZE

EXCESSIVE ERROR MOVABLE TAIL ADDED COMPLEXITY SPAN COMPROMISED AIR TRANS EXCESSIVE VIBRATION
Fig. 14. YUHdl required design changes.

ROTOR HEIGHT

ROTOR MAST INCREASED 24 INS FOR PROD

TOPIG
ROTOR DIAMETER AIRFOIL SECTION ROTOR HEIGHT MAST ANGLE MlSC FAIRINGS BLADE FLAPPING

FLT TEST PROBLEM

PROPOSED DESIGN CHANGE

BLADE TIP SHAPE

WEIGHT INCREASE INCREASE DIAMETER BY 6 INS WADV BLADE TUCK REDUCED THICKNESS OF OUTBOARD 25% TAIL BOOM CLEARANCE INCREASE MAST BY 18 INS FOR SLOPE LANDING ADDED 1.2 DEG FWD TILT .LOW CRUISE SPEED ADDED FAIRINGS FOR BLADE GRIP & HUB *LOW CRUISE SPEED SPRING TORQUE TUBE .INCREASED ALLOWABLE FLAPPING .INCREASE MANEUVER ENVELOPE ELASTOMERIC FLAPPING STOP FOR ADDED CONTROL POWER ABOVE 4 DEG FLAPPING .LOW CRUISE SPEED PLANFORM MODIFIED

REDUCE TAIL POWER SWEPT TIP ADDED T/R BLADE SHAPE *INCREASED COLLECTIVE PITCH AVAILABLE T/R THRUST AVAIL *SIDEWARD FLIGHT CAPABILITY AND ADDED COUNTER WEIGHTS T/R HUB LOW CRUISE SPEED ADDED HUB FAIRING WING CHORD REDUCE FUS DOWNLOAD 6 INCH CHORD EXT TO CREATE A FLAP IN FWD FLT FOR ADDED LIFT HORIZONTAL TAIL POOR FLYING UPPER HORIZONTAL CHANGED TO PROVIDE QUALITIES A 25% CHORD CONTROLLABLE ELEVATOR MlSC FAIRINGS LOW CRUISE SPEED ADDED WING FILLETS MODIFIED MAIN LANDING GEAR FARllNGS REDESIGNED ENGINE COWL TO REDUCE FRONTAL AREA & FLOW SEPARATION
Fig. 15. YAH-63 required design changes (all items only proposed for production).

JANUARY 1990

.FRT TEST PROBLEM


.(I) CANOPY INTERFERENCE m(2J CONTROL POWER FOR .. SLOPE LANDING B L A D E CONTACT DURING PUSH-OVER H I G H CHORDWISE LOADS UNACCEPTABLY HIGH SPEED VIE CONTROL LOADS INADEQUATE LOW SPEED CONTROL UNUSUAL ATTITUDE IN LOW SPEED .INADEQUATE LOW SPEED
CnNTRnl

1989 NIKOLSKY LECTURE

15

DESUGN CHANGE

.ROTOR

HEIGHT

DROOP STOPS LEAD-LAG DAMPERS .MAIN ROTOR BLADES EMPENNAGE CONFIG

. .

TAIL ROTOR SIZE VERTICAL FIN SIZE

. . . .

EXTENDED MAST 10 INS ADDITIONAL 6 INS EXTENSION OPEN STOP ONE DEG W/ FULL RAISED ROTOR REDUCED DAMPER STIFFNESS 20% SWEPT TIP TO REDUCE NOSE-DOWN PITCHING MOMENT FINAL CONFIG ADAPTED AUTOMATIC MOVING STABILATOR MOUNTED LOW ON TAIL BOOM T-TAIL TRIED FOR A SIMPLE FIXED APPROACH INCREASED DIAMETER FROM 100 TO 110 INS REDUCED FIN AREA FIXED FOR SIMPLIFICATION/COST REDUCTION DOUBLED PITCH AXIS AUTHORITY (10% TO 20%) REDESIGN FOR INCREASED STRENGTH & TO CHANGENATURAL FREQUENCY REDESIGN FOR INCREASED STRENGTH INCORPORATED SLOT TO REDUCE NATURAL FREQUENCY RELOCATED FROM UNDER COCKPIT AREA TO TAIL ROTOR VICINITY REDUCED FROM 875 TO 625 SHOTS/MIN ADDED COLLECTIVE RATE SIGNAL TO COLLECTIVE DISPLACEMENT AS ENGINE CONTROL OUTPUT
~ ~

.MOVABLE WING FLAPS DIGITAL AUTOMATIC STABILIZATION EOPT .XMSN SUPPORT STRUCTURE POWER CONTROLS & ACTUATORS TAIL BOOM TUNING TAIL ROTOR ACTUATORS a AREA WEAPON FIRING RATE ROTOR SPEED CONTROL

INTERFERENCE IN RIGHT SIDEWARD FLT EFFECT NOT APPARENT PITCH-UP IN HIGH SPEED FLT H I G H FLT LOADS HIGH FLT LOADS

.
115

EXCESSIVE VIBRATION IN SOME CONDITIONS SENSITIVITY TO FLT VIES EXCITED FUS VIBRATIONS ROTOR DROOP W/ COLLECTIVE PULL

Fig. 16. YAH-64Arequired design changes.

els, quarter scale beingthe largest. Their efforts are summarized below: comlnon objectives YUH-MI Y U R ~ I YAH-63 Y A H . ~ ~~~l
~ a r i c i Vehicle Drag ~ r ~eripn Acro. Airloads.

1110

14

117

Enlpnnage 14 1 15 1 116 117 lralated Tail Rotor ~hrust &Stability Full FUII FUII None ( ~ o t e AII cantractors perrormed drag reduction tests a r m Is1 nigh9 :

In addition, some unique tests were performed. For the record, they are listed here. YUH-60: YUH-61 Rotor Tip Geometry, Maimail Rotor Interaction, and Impact of the IR Suppressor Model Rotor for ForcesIMoments (10 ft. dia.), Tail Rotor AeroelasticStahility (multi-configs),115 Scale Ballistically Damaged Rotor, Engine Exhaust ReIngestion, VibrationIShuffle w1Raised Rotor QuadricycleLanding Gear, Airfoil Tests (both pre& post-flight), Aft Nacelle Area, Gun Fairings, and Main Rotor Hub Canopy Configurations, WingIFlap & Weapons Pylon, Empennage Environment, and Impact of thc IR Suppressor

ing, particularly when rotating components are utilized. The industry therefore has not relied to a great degree on such testing to design acceptable configurations. Some historical comparisons of the wind tunnel occupancy test hours for rotary wing aircraft with recent fighter aircraft developments and the long history of commercial jet airplane developments are shown in Fig. 19. The only exception to this trend has been the recent development of the V-22 (not shown on the figure). Its design nrocess included a~oroximatelv10,000 test hours of mdnv tvws bf different wind'iunnels. 1n.thc 'case of the Boeing 737ij67 designs, the wind tunnel testing accomplished prior to the decision to proceed with the development of the aircraft approximated 10,000 hours each, which is more than double that accomplished in any previous complete helicopter program. The bottom line appears to me to be: Tlieprobability of reduci~lg
the risk o rhe full scale developmerlt o a new rororcrafr car! f f be sig~zificanrlyreduced through more fhoro~rghnind runrzel tesritzg in rlze preliminary design plzase so that the probability

YAH-63 YAH-64

The use of the wind tunnel did play an imponant role in modification to the engine exhaust system and empennage configuratinn of the YAH-64. The differences can be seen bv comparing the photographs of wind tunnel test models presented in Fig. 18. The testing of rotorcraft models in a wind tunnel is significantly more complicated than fixed wing aircraft model test~

of design changes after first flight is minimized. This would follow the lead of the fixed wing industry. The future is already a little brighter. Some wind tunnel testing is being conducted during LHX DEMIVAL to optimize and confirm aerodynamic design parameters. LHX wind tunnel testine will utilize the aoorouriate scale and full-size wind tunnel midels to inve~ti~ate'individual performancelaerodynamicl aemelastic interactions, Amy RFP guidance indicated that these models are expected to include: (1) Small scale (I0 percent to 20 percent) airframe aerodynamic models (2) Medium-scale (30 percent to 40 percent) rotorlairframe interaction models

16

C.C. CRAWFORD, JR.

JOURNAL OF THE AMERICAN HELICOPTER SOCIETY

UTTAS RAISED ROTORS


YUH-60 YUH-61

ORIGINAL ROTOR / EMPENNAGE

COMPETITION TEST CONFIG

FINAL ROTOR / EMPENNAGE

PROPOSED FOR PRODUCTION

Fig. 17. UTTAS raised rotors.

TYPICAL YAH-64 WIND TUNNEL TESTING

ORIGINAL ENGINE EXHAUST

IR SUPPRESSOR

FUS WITH T TAIL

FINAL EMPENNAGE CONFIG

Fig. 18. Typical YAH-64 Hind tunncl testing.

JANUARY 1990

1989 NIKOLSKY LECTURE

J
ROTORWING FIXED WlNG MILITARY 0 FIXED WlNG COMMERCIAL

Fig. 19. Comparison of wind tunnel test times.

(3) Isolated large-scale main rotor models (4) Large-scale anti-torque models
LHX wind tunnel testing will measure airframe drag, stability parameters, and determine parasite power requirements. Wind tunnel testine will also include determination of main and tail nltor elfic~enc~e\, power rcqutrcnlcnts. ~n;~xi~liuln thmst canahilltv. ~ r o ~ u l s i force. and hladc l(rads~bcndincl vc moments. ... Other areas to be evaluated are propulsion system losses, rotor system stability boundaries aerodynamic stability, and aerodynamic interactions. LHX wind tunnel testing has been formulated to aid the design process, and substantially improve design aerodynamic loading problems that have plagued recent helicopter development programs. There may be a qucstion regarding adequate time to evaluate model modifications for optimization purposes, however, because of the sensitive competitive nature of these tests, no information on their exact scope has been made public.

I
I

Weight Estimating Realism


The history of the aircraft industry has been plagued with overly optimistic weight predictions and their adverse impact . . . . on cost, schcdulc, and pcrti~rmancc.Rotorcraft are no exception to fhlsproblcm. It is equally difiercnt tor hclth [he ~ndustrlal . . supplier a i d the military customer to handle weight estimates of new designs that are based totally on parametric analysis. This tends to stifle cost-effective weight saving features. On the other hand, when the initial estimate is based on a bottomsup analysis, its preparation involves an enormous number of man-hours and its accuracy is naturally subject to great optimism. The optimism in the weight estimating process is driven by two almost unavoidable reasons. The first is the desire to win. Notwithstanding the desired preservation of the company, both the prestige of being a military supplier and the patriotism aspects of providing superior equipment to the DoD, fuel this desire. The second reason is an almost blinding faith in new technology. Most new technology efforts, explored eitherthrough the government's Tech Base programs or through the Inde-

I
I

i
I

pendent Research and Developmental activities of industry, have either weight savings or weight avoidance as one major goal. These goals arc often achieved on an individual project basis however, collectively they never seem to sum up directly. I guess I'm saying that the impact of the growth factor concept when adding weight to a design is not as straightforward as when taking weight out. Also, if it is known that the military customer has some target weight objective, no matter how optimistic this objective may be, it is extremely difficult for either the supplier or the buyer to admit that this goal is not in hand. The actual empty weight of any new aircraft is influenced by several other important factors. Changes in customers' requirements during the development process are quite common. Numerous design charges to insure a viable air item usually add significant amounts of weight. Very few "fixes" for reliability, maintainability, aerodynamiclthermodynamics, structural or other mechanical problems result in weight decreases. The cost implications of proposed weight saving features in the original design come to the surface as unfavorable. Techniques like the chemical milling of aircraft bulkheads (OH-6A) to save a few ounces are abandoned in the real world, because the expense of chemical milling does not make it cost-effective. During the preliminary design phase, many trade-offs are accepted which in the real world of flying hardware are quickly found to be "flaky" and thus abandoned. This is particularly true when original trade-off, such as the lack of a subsystem redundancy or reduced design criteria are found to impact safety or to adversely affect maintainability. Finally, the biggest deterrent to achieving a target empty weight is an unrealistic developmental schedule which does not allow time for redesign for the purpose of weight reduction. When low weight is of paramount importance, it must be recognized that such designs are an iterative and time consuming process, requiring schedule allowances. I would like to substantiate these arguments by presenting comparisons of the weight history of the U P A S and AAH designs. The Army Helicopter Improvement Program (AHIP), which has a military designation OH-58D, is also included. Figure 20 presents the percentage change between the initial

18

C.C. CRAWFORD. JR.

JOURNAL OF THE AMERICAN HELICOPTER SOCIETY

DEVELOP SPEC

ACTUAL @ GOVT EVAL

FIRST PRODUCTION ACTUAL

Fig. 20. Realities of weight prediction (empty weight). developmental proposal offered by each contractor for their aircraft relative to the: ncg<~t~atcd dcvclop~nc~it for a) spccification com~nit~ncnts h) actual wclghts :tt the time 111 ll~c official guvcrnment evaluation c) actual weights of the first production aircraft Since the UH-61 and the AH-63 were not fly-off winners, producl~on wcight data ncvcr existcd. 'fhe reason for u s ~ l ~ g the ~nitial proposal cstimatc of thc prin~c manufdclurcr a> the COIIImon denominator for the percentages shown is quite important. Because these initial weight estimates were not generally known to the public, their values are included in Fig. 20. Notwithstanding the government's desire to acquire the lightest weight weapon system practical to meet mission requirements, the government's formal evaluation process contributes significantly to the improvement of realism in empty weight prcdiclions. Government analyses of the group weight statement in great detail, looking for omitted items and unrealistic projections, results in an increase in the full specification weight of about 4 to 5 percent, through the process of negotiating the development contract. Thcsc more realistic weight estimates, in conjunction with less optimism in the aerodynamic predictions of lift and drag and a more realistic assessment of powerplant installation losses, sometimes result in an increase in rotor diametcr before the detail design process begins. While this enlarged rotor is always heavier, it still results in a higher performing rotorcraft. A rotor diameter increase during the negotiation of the developmental contract with the Arn~y actually occurred for both the YUH-60 and YAH-64. Ironically enough, they were the "fly-off" winners, having the best flight performance. Figure 21 is the identical presentation of data in terms of the percentage change from the initial proposal for primary mission gross weights. These percentages also include the adverse impact of higher-than-estimated fuel consumptions, where appropriate. It can be seen that the overall optimism in estimates of the primary mission gross wcight ranges from 8 to 16 percent. Reduction in the mission weight of the production aircraft only occurred for the UH-60. This resulted from a tremendous emphasis on weight reduction between the prototype and initial production aircraft. To enhance their probability of winning the production contract, Sikorsky proposed to eliminate approximately 260 pounds of weight from the prototype. Thcy actually achieved an additional 536 pound weight savings, such that early production aircraft weighed approximately 800 pounds lcss than their development specification commitment. This gallant effort did not last. The nominal weight increase of each production aircraft was almost I pound per ship, resulting in the 980th Army aircraft being 866 pounds heavicr than the first production article. This increased weight is after the deletion of the aft cabin absorber, at 67 pounds, which was in the early production aircraft. Updates to the production specification for approved engineering change proposals result in reccnt aircraft being 35 pounds below the spccification empty weight (I 1,288 pounds). There are many viable reasons for this weight growth pattern; some customer-driven, some for production reasons, some to enhance safety, and many to improve readiness. The compelling point is that weiglzt growth is arl inevitable fact of life.

Power Margins
The most practical way to avoid adversities in mission capability (as a result of overly optimistic empty weight estimates for paper proposals andlor aerodynamic efficiencies that are never quite achieved) is to design for high power margins from the outset. This approach is facilitated by the fact that engine power growth is also a fact of llfe. The succcssful growth of engines is well known history; and, when accomplished in a preplanncd manner, is most cost effective. At retirement from military inventories, rotorcraft engines have experienced a nominal growth of at least 30 percent from their original pre-

JANUARY I990

1989 NlKOLSKY LECTURE

19

INITIAL P R O B l L EST!MTE

15500

14600

14560

126 0 .'90

34 . 7 9

DEVELOP SPEC

ACTUAL @ GOVT EVAL

I7 FIRST PRODUCTIONACTUAL

Fig. 21. Realities of weight prediction (primary mission weight), liminary flight ratings. A specific requirement of the Army's latest T800 engine devclopmcnt planned after LHX was a growth plan for 50 percent over a 10-year period. Both competing teams offer a 10 to 20 percent growth at the outset with vcry minimal change from their preliminary flight rating configurations. Why not use this capability? An equal success story has been the relatively simple growth of rotorcraft transmission ratings. The term "simple" is used here in the context of minimal hardware changcs. The slope of a parametric transmission weight versus horsepowcr curve will predict a weight increase for a given horsepower increase which is at least two timcs the actual weight increase needed to qualify a successful fielded transmission at thc same higher rating. This should make the use of the planned growth of the T800 engine a simple matter if the initial transmission design criteria are not compromiscd. T o quantify the need for a high powcr margin in the initial design, the optimizations in the aerodynamic and wcight predictions will be reviewed for the five designs previously discussed in terms of needed horsepower. Figure 22 illustrates an average power deficicncy of 5 pcrcent due to optimistic hover power required estimates and a 7-15 perccnt powcr deficiency due to unrealistic weight predictions. These factors arc combined in Fig. 23. Minimum and maximum boundaries for useful power margins (excluding vertical rate of climb considerations) are established in Fig. 24. A 15-20 percent power margin is a recommended "insurance policy" to assurc that critical military flight performance characteristics are achieved without compromising reliability, maintainability, or fleet readiness by scrimping on design fcaturcs (solely to minimize empty weight) which enhance the usefulness of the aircraft.

i I

Improvements in Analytical Process


Rotorcraft analysis codes have continuously improved since the dcvelopment of the UTTAS and AAH. Currcnt Rotorcraft

Analysis Codes include C 8 l (Refs. 8 & 9), C60 (Rcf. 20), REXOR (Ref. 21). DART (Ref. 22), Normal Modes (Ref. 23). RDYNE (Ref. 24) and RACAP (Ref. 25). These codes have addressed a broad range of user needs in the various technical disciplincs. However, limitations of current analysis codes include the load of internal interdisciplinary mathematical model consistcncy, inadequate modeling flexibility, theoretical and algorithmic documentation, and havc not bcen cxtensivcly validatcd. For example, these first generation programs have exhibited large differences in predicting helicopter loads. These differences are largely a result of the lack of understanding of the uniquc aeromcchanical complcxitics of rotorcraft. Validation thrusts include data base initiatives for code correlation such as the DNW Wind Tunnel Tests of the Boeing 360, MDHC HARP, and UH-60A rotors, as well as the NASA-Army Black Hawk Flight Test Measurement Program. lnitiativcs in providing comprehensive rotorcraft analysis codes have resulted in programs such as DYSCO (Ref. 26). CAMRAD (Ref. 27 & 28). GRASP (Ref. 29), COPTER (Ref. 30), and 2GCHAS (Ref. 31). GRASP and CAMRAD have improved the state-of-the-art in nonlinear beam dynamic modcling and aerodynamic modcling for rotor air load prediction for use in an integrated comprehensive analysis. However, DYSCO and 2GCHAS provide the most potential in providing a flexible software platform for the evaluation of new technology elements. An exccutive control system provides standard module interactions, run time execution control, a user friendly interface, configuration modeling flexibility and a standard softwarc architecture. This provides a sound basis for future enhancements. While the DYSCO and 2GCHAS codes provide a basis for future enhancements, they currently do not provide quantum leaps in module technology. Additional theoretical research is required in the fundamental disciplincs to achievc the ultimate goal. Only a better understanding of the structural and fluid

20

C.C. CRAWFORD, JR.

JOURNAL OF THE AMERICAN HELICOPTER SOCIETY

25%

5% 0%

I .I :

20% -15% .10% --

I,.:I
Fig. 22. Power margin deficiency. a) Due to missing hover power (tap) b) due to weight growth (bottom)

Fig. 23. Power margin deficiency (due to missing hover power and weight growth).

JANUARY 1990

25%

1989 NIKOLSKY LECTURE

21

0
W E R D 20%

E F
I

15%

M
A

10% I T
5%

R G

I
N 0% DUE TO HOVER PREDICTION
Fig. 24.

DUE TO WEIGHT GROWTH

COMBINED EFFECT

Power margin deficiency (excludes VROC considerations).

mechanical phenomena, their associated engineering representations, and'interactions in an integrated ~ & ~ r e h e i s i < anale ysis will lead to a flexible tool which may be used throughout ihe phases of rotorcraft design, analysis,-and test

Conclusions
The lack of realistic and detailed rotorcraft analytical tools has resulted in many configuration changes being required after the rotorcraft has entered flight test, in order to have a viable weapon system. These analyses short-falls are most prevalent in the areas of estimates for rotor vibration loads, and flight control system loads, overall vibratory levels, as well as.aerodvnamic caoabilitv. The multi-discioline asoects of rotorcraft dynamics significantly complicates the construction of approoriate analvtical tools. The inabilitv of these tools to facilitate design wilich would fly adequateiy "off the drawing board" was illustrated in a development of the Army's competing UTTASIAAH designs. Extensive wind tunnel testina, as a predesian effort, is needed to support future rotorcraft de&lopmknts. while time consuming in the uo-front phase. this testing should minimize changes aitur ftrst fl~ght.Such changes hrr. extret~telycostly and oltcn rc~ult lcss than o ~ t i n i ~f~xcs. in m Malor IixcJ-xinr wind ti~nncl efforts have averaged 12,000 to 20;000 hours oiwind tunnel occupancy time for fighters and 20,000 to 30,000 hours of testing for commercial jet transports. The UTTASIAAH design test efforts ranged from just under 1,000 to slightly over 4,000 hours of tunnel occupancy time. History suggests that the optimism in empty weight and aerodynamic prediction for initial rotorcraft designs warrants a design power margin of 15 to 20 percent. Optimistic weight estimates are driven both by the dcsirc to win and by a blinding faith in new technology. Fixes for reliability, maintainability, aerodynamic, and structural or technical problems almost always rcsult in weight increases. Weight growth is also often oromoted bv chanees in customer retluirements. The validitv , of some initial trade-offs, designed to savc weight, are found not to be in the best interest of safety and readiness. Optimistic

aerodynamic predictions are energized by the same trends. Rotorcraft drive svstems are oarticularlv accommodatine in adaoting dcsign power rnargtns to dctuitl in-scrvicc power utiliz~tiun. A rcvitalirdtic~nof the work < t i huildine and validatinc improved rotorcraft methods is much necdcd. Within the government. this effort has been spearheaded bv the develooment of the Amy's Second Genekition Comp;ehensive ~ c l i c o ~ t c r Analvsis Svstem (2GCHAS). Although this svstem provides an iniegratid framework forthc multi->isciplined natire of the rotorcraft analvsis, development of the individual technical modules to be hsed in ~ G C H A requires extensive researchin S their own right. The completion of this comprehensive analytical system has been plagued by the scheduledelays normally associated with funding problems due to the lack of understanding of its importance within the highest level of the military. In addition, the DoD, its military services, and the industry have all been hampered by their inability to convince congressional budget personnel that the technical complexities and risk of developing new rotorcraft are substantially greater than for their fixed-winged counterparts. Since-there is no "quick fix" to the shortcomings of the rotorcraft analytical tools, the surest way to select the best available weapon svstem is through the "comoetitive flv-off" process. The & .$ s Light observation ~ e l i c o ~ t~ rro & m of e the 1960's resulted in the world's most advanced small helicopter. The UTTASIAAH competitive fly-offs ultimately resulted in successful weapon svstems. Althouah all specification . . requirements are never met, with a fly-off the impact of tradeoffs can be evaluated through the performance of actual missions. The technical efforts of weight, flight performance, structural integrity, and the effectiveness of systems integration can be based on measurements not estimates. After all, competition fly-offs are the ultimate challenge to the industry.

Acknowledgments
This lecture could not have been prepared without the assistance of many members of the Society, who are employed by industry. It contains much historical information that is not

Potrebbero piacerti anche