Compilation of Papers Summarizing Some Recent NASA Research on Manned Military Aircraft

• f

<

< Z

MEMORANDUM TECHNICAL

COMPILATION OF PAPERS SUMMAI_ZING SOME _RECENT NASA __H ON __ _A_ AIRCRAFT By Staff of the NASA

,..N 67- 3307 0

3 "3 _79 N67"

o --- I _'A,q.l[ _,_ IC .O TO 6_I _''_ " _ --,"_ tNA_A CI% OR TVX OR AD _4tJ._BFHI (CAT[GORY) " .,;,v; LTR. NASA, * "_ """'_ •U. _...

_ee _ . , .., 'JL'_l ima_ --" "" III I ............ b ....

t_ _'" -" J.l[--_

_110N

/NATIONAL _UnCS AND SPACE

i .

-I' • " , . -. _,_ .,";_, ._,.:;?_i ..,- . _ _-_ p',,_ G • ,,. ?_,,_- 8 a • • • • • • • • | 6 ' ' • , s I • • ¢ : • ¢ • .I eo • 9eclass_f'ied _y : [|otices i_°'-k_%---- .... -_ float " O n, ChS:Ib9 NATIO_L_L AERONAUTICS #kT) SPACE ADNI_ISTRATIO;: TECHNICAL MEMORA_NDUM X-420 FOHEWORD The nine papers included in this volume were pres.-nted by NASA research staff members at the National Meeting on The _ture of Manned Military Aircraft sponsored by the Institute of the Aeronautical Sciences and held in San Diego, Calif., from August i to August 5_ 1960.

Because, together, they provide a comprehensive summary ol research by NASA in the areas covered, they have been compiled in thic publication for the information of personnel concerned w]-_h the design _ pr,Jcure- ment of manned military aircraft.

i PRECEDING I_A,.,,: ',3L!'J'-;._. I,,(,. t-;L..

-::.-:-.

:.. :.- .-.

::" :: !.: i:. "..

I. _0L AIRCRAFT - STATE OF THE ART . . .

lJ By Robert H. Kirby ...........

II. AERODYNAMIC RESEARCH RELATIVE TO VARIABLE-_ _ION AIRCRAFT . . . By Edward C. Polhamus and Alexander D. Hammond ...................

IIl. SUPERSONIC CRUISE AIRCRAFT . . . By Donald D. Bumls, Cornelius Driver, and Owen G. Morris ........ • . . .

IV. AIR-BREATE[NG PROPULSION SYSTEMS FOR SUPERSONIC AIRCRAFT . .

By Lowell E. Hasel, Willard E. Foss, Jr., and David N.

57 _/ Bowditch ..........................

V. T_E X-15 FLIGET RESEARCH PROGRAM IN RELATION TO THE DEVEIDPMENT OF AIY_ANCED MILITARY AIR_ . . .

By Jack Fischel .....................

VI. HYPERSONIC CRUISE VEHICLES . . . By John M. Swlhart and John R. Henry ......................

VII. SOME STRIETURAL AND MATERIAIS CONSIDERATIONS FOR MILITARY AIRCRAFT . . . By Eldon E. Mathauser, Richard A.

zo7

Pride, and Avrahsm Berkovits ...............

VIII. NOISE CONSIDERATIONS FOR FU_WJRE MANNED AIRCRAFT . . .

IX. SD_IATIC_ RE_ FOR THE DEVELOPMenT OF ADVANCED MASRED MILITARY AI_C_VT . . . By Euclid C. Holl_man and Melvin Sa_off ......................

% £ D=_r'_niNG PAGE BLA,_II¢ blOT FII.WaEI:::;'.

m _..._,.v,,- ......

ill COMPILATION OF PAPERS SL_@IARIZING SOME RECENT .NASA RESEARCH ON MANNED MILITARY AIRCRAFT* By Staff of the NASA I. VTOL kVRCRAFT - STATE OF THE ART By Robert H. Kirby Langley Research Center INTRODUCTION During the last few years a large number and wlde variety of VIOL aircraft types have been studied; some of them promising, some not so promising. The purpose of this paper is to summarize briefly the state of the art in this field and to indicate which of the types are most promising, to bring out applications where they are best suited and to indicate what is needed in the way of additional research and development.

DISCUSSION Before the different types are considered individually, the factors that determine their logical areas of application are discussed.

A basic relationship that exists between the four propulsion types in hovering is shown in figure 1. Hovering effectiveness, which is defined as the amount of vertical lift produced by a given amount of .Dower, is shown as a function of slipstream or Jet velocity. The heli- copter rotor moves a large mass of air downward at a low velocitywhereas the turbojet accelerates a small mass of air to very high velocities.

A good indication of the meaning of hovering effectiveness can be obtained from the fuel consumption of the different types. For hovering with a given payload, the propeller FIDLwill use about 3 to 4 times the fvel used by the hellcopterwhereas the Jet would use in the neighborhood of 25 times as much as the helicopter. Obviously, the hovering time of these higher performsnce aircraft has to be kept to a mlnlmum.

The differences in the slipstream velocities shown ln figure late often cited as reasons for accepting or reJectingvariousVTOL configura- tions. It is probably true that the higher velocities associated wlth Title/Unclassified. -.-

I

| a, @e @ • @ • t Q ¢08 e, propeller, ducted-fau, and turbojet aircraft will increase the severity of the problems of ground erosion and recirculation of dust and debris that has been experienced with helicopters when operating from unprepared bases. Just how much more of a problem this will be, however, seems to be open to question.

Recent National Aeronautics and Space Administration downwash studies with models have indicated that good sod will not d_integrate under the impact of heavily loaded propeller slipstreams. Experience with the Short S.C. 1 in Ireland, has indicated that even turbojet- lifted airplanes can perform certain limited operations from substantial sod.

For logistic support or assault transport missions with propeller- driven aircraft, it would seem to be possible in most cases to find a grass field or a very hard dirt surface for the operation of VTOL air- craft and thereby avoid serious ground erosion problems. This problem is a localized problem right beneath the aircraft. In the surrounding area, the concern that high-velocity slipstr_.ams will be more prone to blow over personnel and equipment some distance from the aircraft is contrary to existing experimental evidence, the reason being the very rapid dissipation of the energy in the smaller, higher velocity slip- streams. (See fig. 2. ) Shown at the top of figure 2 is a helicopter and a typical four-propeller tilt wing, both .weighing the same (about 30,000 pounds). The helicopter has a rotor diameter of 72"feet and there- fore a disk loading of about 7-5- The propellers are 15 feet in dism_ter with a disk loading of 43 or about _2 times that of the helicopter.

The sketches at the top of figure 2 show the rotor and propeller slip- streams as they flow down and then out along the ground. These velocities alon6 the ground are plotted in figure 2 against distance out from the center of the aircraft and the plots show that the velocities decay as the distance increases. The solid line is the horizontal velocity of the propeller slipstream and the dashed line is for the helicopter.

This plot shows that, although the propeller velocities are higher at the start, they decay so rapidly that in a very short distance 2 in this case about 18 feet from the tips, the velocity is less for the propeller than for the large rotor. It would not be expected that personnel or equipment would often be any closer than this crossover point to either of the machines during take-off or landis. It should also be .pointed out that the slipstream fr_ the small propeller is only about _ne-fourth as deep as that for the rotor. This discussion of groun_ effects is not intended to i_ly that the higher slipstream velocities will not cause operational problems in the field; certainly, _ound erosion and related subjects are much in need of research and operational ev_uation, but it does seem that the slipstream problem may not be as great as it is sometimes picturecl to be.

*4r ... ....• - . " • .... : :-. : .-- :..

" .... ::" ::" ": 5 @ • • • • • • 4qt _• As indicated earl_er, the large rotor is the most effective way of producing certical lift but this is only one part of the answer. When the mission calls for higher speed and longer range, an efficient for- ward flight system- something llke that of the conventional airplane - is needed. The power-required curves of figure _ show the shortcomings of the helicopter in this regard. There is a rapid increase in power required for the helicopter at low forward speeds. With some of the higher performance VTOL aircraft, such as propeller tilt-wing, ducted- fan, and turbojet types, the efficiency of the conventional airplane in cruising flight can be approached.

The following areas of application shown in figure 4 where hovering time is plotted against cruising speed were thus obtained: helicopter," for long-hovering and low-speed missions, other roto" types for a little higher speed and range, propeller and ducted-fan VTOL aircraft where higher speed and range is a big factor, and turbojets where speed is the primary consideration.

These _ypes will now be considered individually to determine the most promising ones in each area.

The helicopter, of course, is already well established and, with expected improvements, should continue to be the best VTOL for missions such as flying cranes, rescue, and forward-area operations where long hovering time is required and where low speed and short range are acceptable.

Figare 5 shows two other rotor types: the compound or dual pro- pulsion on the left and the tilt rotor on the right. Here the disk loadlngs have increased only a little above the helicopter, about 8 or so on the compound and up to about 20 for the tilt rotor; thus, these are promising types for applications where a large amount of hovering and low-speed operation is still required but where somewhat higher cruising speed and range are needed than can be achieved wlth the helicopter.

The state of the art is further advanced for the compound hell- copter than for the tilt-rotor machine, mainly because its development has been actively pursued in England as the Rotodyne. It appears that a machine of the Rotodyne type may operate reasonably satisfactorily as a commercial transport provided certain operating problems, such as thin high noise level and high operating cost, can be solved. The development of a military version of this aircraft would seem to be a fairly straightforward procedure.

The technical feasibility of the tilt rotor has been demonstrated successfully. A few objectionable handling qualities have been discovered but they seem to be largely functions of t_le very lightly loaded rotors ! J o , ee • ee • @o o @ t0o • • • • @ @@ of the research machines and might not be such serious problems on a trs_sport-size aircraft with more heavily loaded rotors. Although it does appea_ possible to develop an operational tilt-rotor machine, interest in this type seems to be limited because of strong c_npetitlon from two other VTOL types, the compound helicopter which could be avail- able for o_erational use sooner and propeller configurations which offer better cruising performance.

Propeller-driven VTOL's (particularly tilting ring and deflected slipstream types) have received much attention. Figure 6 illustrates a configuration that c_bines the two types somewhat by having a tilting wing with a large chord slotted flap and results in a configuration that is considered to be one of the most promising V_L types, particularly, for use on short or medium transport missions. A configuration like this would take off and land in a short distance (STOL) where possible, such as from rear supply areas and for ferry hops, but would have the vertical take-Dff and landing (VTOL) cs_pability where needed. For one particular mission, that of a small lo6istic support transport which was studied in connection with an ASR (Army Service Requirement) program, a tilt-wing configuration of this type seemed to offer very promising _erformance in terms of payload, range, and operating cost.

Fairly extensive wind-tunnel and flying-model research on advanced configuratior_ such as this type have indicated solutions to most of the peculiarities of the early propeller test beds. It As now felt that there are no technical barriers to the design and construction of a machine of this type for obtaining operational experience. Of course, additional research and development is still needed.

To date, d ucted-fan VTOL aircraft have appeared to be generally less promising than other types but there are three ducted-fan applica- tions for military use that should be mentioned. The first configurations are for a special lov-speed application and have been termed flying plat- forms and light combat aerial vehicles (sometimes called aerial Jeeps).

These machines are the result of an effort to give the Ar_ a utility machine that vould be simple, compact, and easy to fly. Research to date, hoverer, has not revealed any configurations likely to meet the requirement of being simple and easy to fly and the machines tested to date have been restricted in forvard-fllght performance.

The second ducted-fan applications are the fan-in-uing and fan-In- fuselage configurations. These aircraft have d-cted fans buried in the win8 or fuselage for vertical take-off and landi_ and for cruising flisht the fans are covered over and conventional turbojet propulsion is used. Recent research has revealed an unexpectedly_ 8evere m_blem in the transiti_n speed range for bo_h these subaersed fan types and for ltfttns-Jet engine eonf$gurattons. The probles arises from the inter- ference of the fan or _et exhaust vlth the free-streem atrfloe. Figure 7 illustrates the effect of this interference and shovs a planform vith | • .0 0_ • oso • @o@ _ej J e _'0 • • O 6 • ee • •

eJ6 ljp; : : : :'" :'"

high velocity exhausts in the center, either from jets or fans. At the low forward speeds during the transition the exhaust interferes with the free-stream flow, causes it to speed up to get around the exhaust, and also separates in some regions behind the exhaust; as a result nega- tive pressures occur over large areas of the lower surface. The nega- tive pressures are represented here by the shaded portions, the darker shades meaning lower pressures. If it is not te!_en into consideration in designs that have flat areas beside and behind high velocity exhaust, this low-pressure region could result in large losses :n llft, large pitching moments, and stability problems during the transition between hovering and forward flight. The fan-ln-wlng or fa_-in-fuselage types might find military application where high subso,,ic or supersonic speeds are needed, but there are problems that will have to be carefully resolved before the system is r,eady for operational use.

The third and most successful ducted-fan VTOL to da_e is the tilting- duct type shown in figure 8. The wing-tlp-mounted duct_ rotate through 90 ° for hovering and forward flight. The technical feasibility of this type has been demonstrated and its state of the art is approximately the same as tho_o fcr the tilt-rotor and tilt-wing configurations elthough, at this time, not quite as well supported by wind tunnel and flight-test experience. This type could have merit for certain military applica- tions where compactness is deslr_d at the expense of some hovering effi- ciency and short take-off and landing capability.

A number of turbojet VTOL configurations have received attention, such as ones using tilting Jets, deflected Jets, and small lightweight lifting Jets. Altho,_h these types have been demonstrated to bc techni- cally feasible with varying amounts of research and several Jet research aircraft have been successfully tested, no operational aircraft have been flown. There 18 one small operational Jet, howeve that is expected to fly in the near future. It i8 the Hawker P.I127 transonic striko aircraft shown in figure 9.

With a VTOL weight of about 1_,000 pounds the Hawker is powered with a Bristol BE-5_ turbofan engir_ with s__velling nozzles, two on each side of the fuselage. In hovering, the four nozzles are pointed downward and the transition is performed by rotating the nozzles rearward.

In this country most of the interest of the military h.., been In.VTOL supersonic fighters But their development is expected to follow well behind that of other VTOL types. The operational experience to be gained w'J.th this subsonic configuration should provide some of the information needed for proceedi_ to the supersonic applications.

w 6 .... :: :: "::: The plot of hovering time agalnst cruising speed (fig. i0) is again shown to indicate the four most promising types at the present s_ate of the art in their respective areas. Now that the feasibility of these types has been established, the VTOL aircraft field is now ready for the next step, that is to build some of these higher performance machines so that the operational experience needed to determine their potential can be obtained and to define more clearly the service requirements in the various areas. It is expected that this experience will be gained in the near future with the machines on each end of the chart (fig. i0), the compound helicopter and the turbojet fighter. It appears highly desirable, therefore, to obtain as soon as possible an operational tilt- wing machine which fills in the middle of the chart and seems particulsrly well suited for transport missions that require long range and where higher speed is advantageous.

By indicating that operationally useful machines can and should be obtained at the present state of the art, it is not meant that a great deal of research is not still needed in this field. On the contrary, before the full potential of VTOL systems is realized, there is a vast a_ount of researcht development, an_ experience needed.

Problem areas comBon to all of the VTOL types and needing additional research and d2veloi_ent are: cost and weight of airframe and propulsion system, handling qualities, all-weather capability, and ground erosion and recirculation. Cost and weight both must be reduced for the air- frames and, much e_phasls _ust be placed on propulsion systems, including • gearing, rotors, and propellers. Continued research is needed on handling qualities and all-weather capability is something that must be achieved to utilize these machines fully. In addition ground erosion a_i reclrcula- " tion needs evaluation, particularly for the higher performance types.

Even the helicopter, which is being obtained in increasing numbers, is far from an optimum system. Besides needing a drag reduction program to improve its range and endurance, the helicopter has many other areas still needing attention. The cyclic loads, particularly in the cruising range and vibration are problems. All-weather capability should again be emphasized for the helicopter an_ the need for improved behavior when operating in or near severe turbulence.

The compound helicopter will also have some of these problems plus a few of its own such as rotor instabilities at the higher forward spec_Is.

Reduction of hub, pylon, and interference drag will be even more important at these higher speeds. If tlp-drlven rotors are used, the noise level is a big problem ar_, in this connection the question of whether a tip- driven or gear-drlven rotor is best for the c_pound helicopter is not in En_ therefore, their relative _rits are in need of evaluation.

; ! "'" i"" i" ...: ....... . - ... ". -

• • _e, oo • e, a • .. • 4,@ @' fqp t • The tilt-wing configuration needs detailed research on loads and stresses leading c the development of lightweight propellers, research on gearing and turbine _ngines leading to lighter weight and especially lower cost and greater freedom from maintenance, and continued structural research; all of these lead to a higher percentage of useful load to gross weight.

In addition continual ae_rodynamic and flight research on improved performance and handling qualities and operational evaluation of compro- mise factors such as VTOL and STOL capability, speed as opposed to range, and the seriousness of ground erosion problems are needed.

Some of the more important research needs o _ turbojet confi6urations are discussed next. The need fOT the development of lightweight engines cannot be stressed too greatly. Noise and ground erosion and ingestion problems will have to be evaluated and reduced. And finally, operating problems could be especially severe with turbojet configurations.

Research is being carried on in many of these areas at the present time and others will undoubtedly be studied in the near future. Some, such as the development of really lightweight, inexpensive gas turbines, will need concerted effort, time, and ingenuity to solve, and many are such that only experience with us ._ful machines can effectively show the way.

CONULUDING REMARKS Although a great deal of research and development will be required before operational VTOL aircraft will be obtained, the state of the art in this field has advanced to the point where operationally useful machines of some VTOL types can be designed and built. There is a great need now for experience with such aircraft to determine their capabilities under conditions of field operation. Efforts should be made as soon as pos- sible to obtain operational machines to provide this experience.

"- ,_ QO l HOVERING _:_ EFFECTIVENESS '_ : _PROF'BJ_ER ¢ RBOJET I I I I I !

I0 SO I00 300 1,000 3,000 SLIPSTREAM OR JET VELOCITY, MPH Figure l.- Varlatlon of hoverlng effectiveness wlth slipstream or Jet velocity for various propulsion types.

I WT • 50,000 LB DIA.- 72 FT DISK LO_ING -7.5 DIA. - 15 FT LO£OING -4S /_//////////////////////,/_/// / VELOCITY 120 ALONG GROUND, 80 FTISI_ 2_0 40 i | i i 60 80 I00 120 Figure 2.- Comparison of slipstream velee£ty alon_ the ground for a heli- copter and a typical four-propeller tilt-win8 configuration.

LIJ POWER REQUIRED

\\

FORWARD SPEED Figure 3.- ¢ygicat vsrlstion of power required _¢th fo_d speed for various VTOL types.

HEUCOPTERS OTHERROTOR V'rOt.

HOVERING TIME PROPELLER VTOL D_ FAN VTOL TURBOJET VTOL----_ v 0 CRUBNG SPEED Figure _.- LogiceCL areas of applications for various VTOL types.

i0 _J (a) Compo_md helicopter. (b) Tilt-rotor configuration.

Figure 5.- Rotor types.

J Figure 6.- Tlltlng-wlng eonflguratlon.

% o°i • • • .:.. • oa o66 eo FREE-STEAM FLOW !

Figure 7.- Effect of interference between exhaust and free-stres_ flow.

Figure 8.- Tilti_-duct configuration.

_ -- ,_--_.=.q=. ,

I

Q Figure 9.- Hawker P.I127 turbofan configuration.

HEUCOPTERS COMPOUND HELICOPTERS HOVERING TIME PROPELLER TILT-WING TURBOJET FIGHTER 0 CRUISINGSPEED Figure I0.- The four most promising types at the present state of the art in their respective areas of applleation.

) II. AERODYNAMIC RESEARCH RELATIVE TO VARIABLE-SWEEP

N 6 7 - 3 3 0 7 2

By Edward C. Polhamus and Alexander D. Hammond Langley Research Center INTRODUCTION The development of a multimlssion military aircraft would be highly desirable both from the standpoint of easing the strain on the national budget by reducing the number of aircraft types and of providing versa- tility which would increase the effectiveness of the aircraft in the per- formance of a given mission. Some of the capabilities which might be required of such an _ircraft are shown in figure i along with their respective aerodynamic and configura$ion requirements. The first three capabilities are grouped together since they all require good subsonic characteristics. The first, a long loiter capability for combat air patrol and the second, a long ferry range for efficient aircraft deploy- ment both require a high subsonic lift-drag ratio. The third capability, STOL, is desirable for carrier and short-field operation and requires the development of high lift. All three of these capabilities can best ue obtained wlth a high-aspect-ratio wing having a large span and a low- sweep angle. The fourth capability is that of a high-altitude supersonic attack or intercept and requires a high _.ift-drag ratio at supersonic speeds which dictates a rather slender _ ufiguration with a moderate-span wing which is either very thin or highly swept. The fifth capabl._ ity listed in figure 1 is that of a low-altitude hlgh-speed attack that would increase the probability of long-range penetration of antiaircraft defenses. The high dynamic pressures encountered on the deck at high speeds require a iow-llft-curve slope to reduce the gust-induced normal accelerations, _,d low friction and wave drag (drag due to llf_ is inslglliflceat at high dynamic pressures) to assure sufficient speed and range. In order to best satisfy these requirements, a slender aircraft having little or no wing is required. It is apparent from figure 1 that these five capabilities are highly incomtmttble and that an efficient multi_sslon aLrcraft will require a mesas of varying its aerodynamic characteristics. This car, be best accomplished wlth some ty:._ of varlable-wing geometry. There are, of course, several types of variable- wing geometry. However, im vlev Of the extremely large variations in wing span desired, variable viz_ sweep, as indicated In the lower rlght sketch of fisure i, appears to provide the best method. The Langley Research Center of the National Aeronautics and Space Admlnlstratlon has therefore initiated a research progrsm +.o provide the aerod_mic informa- tion needed for the development of a variable-sweep multhaisslon zLtlttary aircraft, mld it _s the purpose of this paper to briefly describe some of the results of _hls program.

SYMBOLS lift coefficient CL lilt-curve slope C_ rolling-moment coefficient pitching-moment coefficient longitudinal-stabillty parameter yaWing-moment coefficient Cn c vlng chord Di induced drag g acceleration dae to gravity

L/D lift-drag ratio

maxlmumlift-drag ratio M Mach number A_ normal-acceleratlon Incr_ent ph/ , nondimenslonal rolling velocity W _mlght of aircraft amgle of attack tall dihedral Pt horlzontal-taildeflectlom • • • • Q If) r • Qe • m 0 ' spoiler-control projection D S A wing sweep angle WING DEVELOPMENT Variable wing sweep, of course, is not a r_w concept m,d the fe_n!- bility of in-flight-sweep variations has been demonstrated with the Bell X-9 and the GrummanXFlOF. Application to a modern multim[ss[on military aircraft, however, requires _dditional considerations. First, current military requirements are such that considerably hitcher sweep an_]ez must be considered, and second, a method of eliminating the need for the fore and aft wlrg translat'on used in the X-9 and F]0F would be highly desirable. This translation was used to control toe stability of the aircraft and consisted of a forward shift of the wing as the sweep increased; thereby, a relatively constant stSoillty margin is provided.

This translation, however, causes additional mechanical coml,lex[ty, dr_[ penalties, and a reduction in usable fuselage volume, each of which will be 15/rther compounded by the higher wing sweep angles currently needed• The first study, therefore, was directed toward the use of large sweep variations and the development of a method of cont_olllng the longitudinal-stabillty variation with sweep that would not require the fore and aft wing translatlol used in the previous aircraft,. Four of the aircraft arrangements studied are shown in figure P with the hi,h-sweep condition show1_by the solid outline and the low sweep, by the dashed outline• A complete description of this study can be found in reference i. The configuration shown in the _lpper left was an all- wing design utilizing an 80 ° arrow planform and pylon-mounted engines on the outer wlngpanel which were used in an attempt to control stabi]ity by means of center-of-gravlty shifts. The arrangement in the upper right was a more conventional englne-in-fuselage configuration utilizing a canard surface for )ongltudlnsl control and aft foldlr_ tails as an aid in the control of stability variations with wing sweep• This wing had a le_ll_-ed_e _weep of 79° for the highly swept condition and had s wing pivot located im close proximity to the fuselage. The design shown in the lower left utilized the same wing; however, a larger aft tall was used, _ canard w_%s removed, and longitudinal control was obtained with elevons for the highly swept conditions _nd with the aft tall for the _nswept conditicu. All three arrs_ements exhiblt_d unde- sirable stability characteristics which are described in reference i.

The design showa in the lower right of figure 2 consisted of a fairly couventlonal effacement having a fixed a_'_ tail; however, the wlng planform was modified to incorporate improvements with regara _o sta- billty that were indicated from the results on the previous conflgura- tions. This w_ had an outboard pivot and a fairly large and effective fixed portion of the wing. With this arramgem¢ot the llft of the outboard i s . • , c 16 " ° 0e _ • ,°• ,Q •, •_ panel, which increases with decreasing sweep, and the llft of the fixed portion co_ine in such a manner as to keep the stabil_ty relatively constant. For this configuration essentially the same stability was obtained in both the high- and low-sweep positions with only minor varia- tions in the intermediate-sweep range. It appears therefore that with this type of variable-sweep wing the need for wing translation can be eliminated. The aerodynamic characteristics of this oonfiguratien throughout a large Mach number range can be found in references 1 to 6.

In order to illustrate the importance of the pivot location and the L amount of fixed area, wind-tunnel tests were made of the two variable- sweep wing arrangements shown in figure 3. The horizontal tail has been omitted in the interest of clarity since the purpose of the drawing is to compare the two wings. However, the same Lorizontal tail was used 7 in conjunction with both wings. The type of variable-sweep wirg just described i_ shown in both the 29 ° and 7_ ° sweep conditions by the solid lines• The distinguishing features of th_s wing are the large fixed area ahead of the pivot and an outboard-pivot location. The large fixed portion provides the aerodynamic solution to the stability variations with sweep mentioned previously while the outboard pivot provides the large span variations desired. Shown by dashed lines is a variable- sweep wing having essentially the same area and sweep conditions, but having an inboard pivot and a small fixed area similar to that of the X-9 wlmg. While the geametry of the two configurations is quite similar in both sweep conditions, extremely large differences in longitudinal stability exist, as shown in figure 4 where the pitchlng-moment coeffi- cient is presented as a function of lift coefficient for the two config- urations at low speed. On the left the results obtained for the inboard pivot are plesented while on the right the results obtained with the outboard-pivot configuration are presented. In the top portion of fig- ure 4 the characteristics of the configuration without the horizontal tall are shown for two wing-sweep positions. For the inboard pivot the results indicate a large stability shift, in the stable d_rection, as the sweep is increased frem 30o to 70 °. It is, of course, this type of variation in longitudinal stability which dictated the use of fore and aft wing translation on the X-9. However, for the outboard-pivot con- figuration which, as it will be recalled, provides approximately the same variations in wing span as the inboard-pivot configuration, the results indicate an actual reduction in stability as the wing was swept from _o to 75 °. This unusual situation illustrates the powerful effect of the fixed portion of the wing in controlling the stability variation with sweep and indicates the possibility of _ctually counteracting the Much number effect on stability. Because of the reduction in lift-curve slope with sweep the tall contribution to stability increases with increasing wing sweep. The t_il-on results are presented in the bottom portion of figure _ and an extremely large increase in s_abilit-y is indicated for the Inboard-piv0t wlmg. This increase when combined with the increase due to Much number would result in excessive stability

ii. i . i •

e • ¢ • • • • • • • i_• e¢ and trim drag at supersonic speeds. For the outooard-plvot wing, hcwcvcr, only a very slight increase in stability occurred with increaslr_ sweep.

It should be noted that an additional lO ° of sweep variation was utilized with the outboard-pivot configuration• These results indicate that by pruperly proportioning the areas of the fJ_ed and ro_ating portions of the wing the stability variation with sweep can be controlled without the need for translation. Complete aerodynamic characteristics of these configurations at subsonic, transonic, and supersonic speeds are presented in references 7 to 9, respectively.

AIRCRAFT CONFIG[_ATIONS STUDIED Since the preliminary study just described indicated that by careful wing deslgn longitudinal stability could be handled throughout a rather large sweep range, a program was initiated to investigate the aerodynamic characte'istics of variable-sweep aircraft designed specifically to meet the mnlt_mission requirements listed in figure 1. A large number of con- figurations were considered and eight wlnd-tunnel models were constructed and tested at subsonic, transonic, and supersonic speeds• Two of these configurations, which illustrate the major configuration considerations, are shown in figures 5 and 6.

These configurations differed somewhat from those described pre- viously in that the wings could be completely folded which, as previously mentioned, is desirable from both performance and gust-acceleration con- siderations in the low-level high-speed-attack phase of the multimisslon.

With regard to perfo_zance th_ fully folded wing, in addition to reducing the friction drag, should allow complete area-ruling benefits on tran- sonic wave drag to be more nearly realized. The method described ia reference lO was utilized in area ruling the configurations. In connec- tion with the gust-lnduced normal accelerations the fully folded wing provides relief through both a reduction in lift-curve slope and an increase in wing loading.

The configurations were designed around two turbofan engines and their volumes were compatible with those of aircraft .In the 60,O00-pound class. The configuration (7) shown in figure 5 is characterized by a wlng-pivot location within the fuselage (see section A-A) and a rela- tively small fixed portion of the wing. While this arrangement exhibits fairly ?.arge increases in longitudinal stability w_th increasing sweep, it allows a large portion of the wing to be hidden for the low-altitude attack and. may afford some structural advantage over configuratioas having the pivot located within the wing. The results are therefore valuable in assessing structural_ performance, and longitudinal-stability i "- ' ' • .* • t • , ,s • _ ,_ , • _p 18 ......

o Q * 4 " - _ ° .i ,iv . o _._ . . q_ trade-cffz. Confi_Jratl%n 8 (see fig. 6) was quite siLilar to config- urrtion 7 with the main differer.ce be-_ng essociated with the outboard- p_.vot locatlon and the relatively large fixed-wi_,g area. _lis configu- ration, while possibly having larger structural penalties because of the pivot location, exhibits mort desi :able iongltu,linal stability characteristics than configuration 7, and would be expected to have lower trim dra Z at supersonic speeds.

Since the most complete data available at the present are that for ,-onfiguration 8 and since it exhibits desirable longitudinal-stability I characteristics, it will be used throughout the remainder of this paper I to illustrate some of the aerodynamic characteristics of variable-sweep multimis_ion aircraft. The wind-tunnel models were 1/24 scale and a photograph of configuration 8 with the wings extended is shown in fig- 7 c ure 7- The jet-engine ir!ets were designed and constructed so as to provide the proper mass flow for a Mash number of 1.2.

PERFORI;_%NCE The effect of wing sweep on the maximum lift-drag ratios at sub- sonlc and supersonic speeds is shown in figure 8. The wing had a stream- wise thickness of 6 percent when swept 25 ° and had a rounded leading edge. The subsonic results are presented for a Mash number of 0.6 and have been corrected %o full-scale turbulent skin-frlction conditions corresponding to _n altitude of 50,000 feet. The supersonic results are presented for a Mash number of 2.2 at an altitude of 60,000 feet• For the supersonic-attack mission it will be noted that a wing-sweep position of approximately 79 ° would be desirable. It will be noted, however, that at subsonic speeds the max.lm_ llft-drag ratio for this sweep would be somewha_ less than i0 and a flxed-wing aircraft wo_[Id therefore have relatively limited loiter and ferry capability. However, for the variable-sweep aircraft with the wings rotated forward to 25 ° sweep, the large increase in wing span increases the maximum llft-drag ratio to slightly in excess of 18 at subsonic speeds. This, of course, would nearly double the endurance time for the loiter mission and would result in a large increase in the ferry range.

In addition to loiter, ferry, and high-altltude supersonic-attack capability, a high-speed low-altltude-attack capability is highly desir- able. Because of the high d_namlc pressure encountered during this mission, the drag due to llft, even for a wingless configuration, is a small po_-tion of the total drag, and the maximum llft-drag ratios became rather meaningless with the minimum d'ag in pounds becoming of prime importance. T-nerefore, in ordor to i_'ustrate _he role that variable sweep plays in connection with the drag in the low-altltude-attack mission, figure 9 has been prepared. Here the total drag associated with a _AJ 60,000-pound airplane in level flight at sea level is presented as a function of wing sweep for various Mach numbers. The resllts indicate that for a supersonic low-altitude attack at a Math n/tuber of 1.2 a considerable reduction in level-flight dr84_ is obtained as the wings are rotated back because of the reduction in wetted area and wave dry.

It is interesting to note that even at a Mach number of 0.9 the benefit of the fl, i!y folded wing is still realized. At a Mach number of 0.6, however, the 8_ies of attack requireG for level flight become large enough so that the drag due to lift becomes significant and the low win6 sweep provides the lower drag. The drag due t,_ lift at M = 1.2 is shown by the .hatched area, and the low value indicates that if additional drag improvements are to be realized for the supersonic mission, reduc- tions in friction and wave dra@ through reductions in wetted and maximum cross-sectional areas must be resorted to.

EFFECTS OF LIFT-CURVE SLOPE Figure i0 shows the variation of the lift-curve slope (based cn a common full-scale area of 600 sq ft) with Mach number for the 25 °, 753, sald fully folded wing-sweep positions. The main point of interest is the fact that in the fully folded condition the lift-curve slope is reduced to 25 or 50 percent of that associated with the fully extended (25 ° ) wing. This large variation in lift-curve slope is of interest mainly in connection with gust and pull-up response (both of which are important for the low-altitude attack), and its effect on these are illustrated in figure Ii.

The conditions represented are for an aircraft weighing 60,000 pounds (corresponding to a wing loading, for the extended case, of about 90) and flying at sea level. Presented on the left of figure iI is the response to a 90 fps sharp-edged gust as a function of Mach number for several wlng-sweep positions - 29 °, 79o, and 109o. The results indicate rather large reductions in the gust response as the wing sweep is increased.

At a Math number of 1.2 with the wings fully folded, the gust response is slightly less than that encountered with the 79 ° wing position at a Mach number of 0.90 and considerably less than that for the 29 ° wing position at a Mach number of 0.6. It should be pointed out that inas- much as a comnon reference area was used for the lift-curve slopes there is no effect of wing loading on the gust response. It should be noted that in the fully folded condition the wing outer panel is locked to the fuselage; therefore, aeroelastlc effects would be expected to be small for both the model and the airplane. For the 79° sweep position, however, aeroelastlc effects would be somewhat greater on the airplane tha11 on the model (aluminum with streem%_ise thickness of 6 percent in 29 ° position and a dynamic pressure of 935 Ib/sq ft at M = 1.2), and some reduction in gust accelerations would be expected.

2O • • • • t • The low-lift-curve slopes which make possible the reductions in gust accelerations raise a question as to the effect of lags (due to the larger rotations required) in pull-up response in connection with terrain clearance during the low-level-attack mission. This effect is shown on the rig_ht-hand portion of figure ll where the longitudinal dis- tance traveled while gaining an altitude of _0 feet following a pull-up from level flight at a Mach number of 1.2 at sea level is sho_m as a function of the steady-state normal-acceleratlon increment achieved in the pull-up. Resalts are presented for the ideal no-lag condition CL_ = _ in which the required rotation is obtained instantaneously and L for three wing-sweep positions. If a steady-state normal-acceleratlon increment lkn of 5g is assumed, it will be noted that even for the ideal case a distance of 1,400 feet would be traveled before _K) feet of alti- tude were gained. The lag associated with the lov-lift-curve slope of the fully folded wing at 5g increases the distance traveled to approxi- mately 2,400 feet, but it will be noted that this is only about lO0 feet or about I/i0 second greater than that experienced with the wing in the 79 ° sweep position and only about 400 feet greater than that with the wing in the fully extended position (A = _o). It therefore appears that large reductions in vertical-gust response can be obtained with the fully folded wing without seriously reducing the pull-u9 response.

LONGITUDINAL STABILITY The effect of Mach number on the longltudlnal-stabillty parameter Cmc L is presented in figure 12 for the vsa-ious wing positions. Shown are the subsonic data with the wings ex_Gended, the transonic data with the wings fully folded, and the supersonic data with the wings in the 79 ° position• There are several items of interest• First, it will be noted that the shift in static margin from M = 0.6 with the wings extended to M = 2.0 with the wings in the 75 ° position is only about 8 percent of the mean aerodynamic chord of the extended wing. It appears, therefore, that by careful wing design large sweep variations can be combin__i with large Mach number increases without encountering stability increases greater than those associated with fixed-wing aircraft. In fact, it appears possible to actually reduce the stability changes.

Secondly, while a considerable reduction in stability occurs at transonic speeds with the wings fully folded, a rather large increase can be produced by the use of 20 ° of negative dihedral in the horizontal tail as indicated by the filled-ln circular symbol. This is due mainly to the favorable effect of the sidewash component of the traili_ vortex induced velocity at offcenter positions. There is an additional benefit from the negative tall dihedral in that it reduces the stability at M = 2.0 (no favorable sidewash component), thereby, a reduction in the supersonic stability shift results. It appears from these results that reasonable stability characteristics can be obtained despite the large Mach number and sweep range that may be required for multimission military aircraft.

LATERAL STABILITY AND CONTROL With regard to static lateral and directiona_ stability it is suf- ficient to point out that the only problem encountezed was the usual one of directional instability at the higher angles of attack at supersonic speeds. It appears, however, that the 20 ° negative dihedral in the hori- zontal tail suggested in connection with the longitudinal stability would be sufficient to take care of the directional problem.

Figure 13 shows the lateral-control characteristics. For wing sweeps up to 75° or 80 ° the lateral control can be obtained by the deflection of a spoiler-slot-deflector control on the movable outboard wing. The control would be located just ahead of a trailing-edge high- lift flap having a full spar, (needed for STOL) and would span the out- board wing. The data on the left of figure 15 were obtained at a Mach number of 0.15 for a spoiler projection of i0 percent of the wing chord.

The deflector projection w_s 5/4 of the spoiler projection. The varia- tion of the rolling-effectiveness parameter pb/RV with angle of attack is shown for the wing swept 29 ° (solid curve) and 7_ O (dashed curve).

The spoiler-slot-deflector contro_ has more effectiveness at low angles of attack when the wing is at 29 ° sweep than for the 79 ° swept wing.

However, the rolling effectiveness is approximately the same at high suagles of attack for both wing sweeps. The ms4_nitude of pb/2V required for fighter-type aircraft at subsonic speeds is in the order of 0.06 to 0.07, and it will be 1,seful to note that with the 25 ° wing position the requirement can be met with a lO-percent projection.

When the wing is fully sweptback for the low-altitude-attackmisslon the spoilers, of course, cannot be used but lateral control caube obtained by differential deflection of the horizontal-tail surfaces.

Data for differential tail deflection have been obtained at a Maeh num- differential ber of 1.2. The rolling effectiveness is shown for ±_v tail deflection on the right of figure 15 and the level of control effec- tiveness for the angles of attack shown are comparable to those obtained at subsonic speed for the spoiler-slot deflector. For this Mach number, however, the roll-rate requirement is greatly exceeded and a control sensitivity problem maybe encountered. Another problem is indicated on the lower part of figure 15 where the variation of the ratio of yawing- mmment coefficient to rolling-moment coefficient is shown for the differ- entially deflected horizontal tall on the right and the spoiler-slot

deflector on the left. The magnitude of the yawing moment due to

rolling moment is considerably higher for the differentlal-tail control than for the spoiler control. This is due to the large induced side- force load on the verLical tall resulting from differential deflection of the horizontal tail. For this reason it is felt that an effort should be made to develop a lateral control on the wing for this maxlmum-sweep case. The most obvious possibility is the use of the leadlng-edge droop- nose flap as a trailing-edge flap-type control when the wing is fully swept as indicated in the sketch by the shaded area, or the use of a split flap-type control as shown by the dashed lines. I_ta are currently being obtained at transonic speeds on both of these lateral-control devices.

OTHER AERODYNAMIC CONSIDk-'RATIONS The wlnd-tunnel studies described previously and in references ii to 14 indicate that with proper design, no extreme static aerodynamic problem areas appear robe associated with the variable-sweep multi- mission aircraft. However, for dynamic conditions, further analyses will be necessary in order to evaluate fully the multlmission capabilities of the variable-sweep aircraft. These analyses should include studies of the results of wind-tunnel investigations already underway including flutter and buffet tests, oscillation tests in pitch audyaw, and tests under conditions of steady rolling, as well as simulator studies of possible roll-coupled divergence, roll-to-sideslip ratio, and control sensitivity.

MLqA_LMISSION PERFORMANCE EST_4ATES So far the discussion has dealt with the aerodynamic character- istics of variable-sweep multlmission aircraft; however, by ws_ of con- clusion some preliminary calculations of the performance capabilities of a possible multlmissionmilitary airplane are given. It should be pointed out that, while these calculations were based on rather rough weight and englne-performance assumptions, it Is felt that they are illustrative of the possible multimission capabilities. The deslgnwas based on configuration 8, had a take-off gross weight of 63,000 pounds, had 28,000 pounds of internal fuel, and was powered by two turbofan engines. The calculations Ind/cated that with the wings extended take- off and land/ngdistances, over a SO-foot obstacle, of less than 5,000 feet were possible with relatively simple high-lift devices. A low-level strike radius of apprcxlmatelySOOnauticalmiles appears possible with half of the outbound leg being acecaplished at a Mach number of 1.2with the wings folded. A strike or intercept radius of --all J e_ ,_ • ,.. • ,,_o -e _. _.. _. : : -" . _ - - ..

0 • ot 2_ ,,. ... .. ..

a I , ! i J |t e e _ eD oo • • 4g t 900 nautical miles could be obtained at a high altitude and a Mach num- ber of 2.5 with the wings in the 75 ° position. By adding 3,500 pounds of fuel in the bomb bay and 8,000 pounds of external fuel, with the wings extended, a ferry distance of approx_.mately 6,000 nautical miles appears fes.slble. From these performance estimates it appears that a variable-sweep aircraft can provld_ for a great deal of versatility and that one aircraft can actually perform several missions efficleutly.

In addition_ the wlnd-tunnel studies described indicate that with proper design no extreme static aerodynamic problem areas appear to be associ- ated with this type of aircraft.

: .: :. ..... . .....

_ t Q • • • oo • 24 "! !! .... .

• • Q • • • • Q'_* Qo REFERENCES I. Afford, William J., Jr., and Henderson, William P. : An Exploratory Investigation of the Low-Speed Aerodynamic Characteristics of Varlable-Wlng-Sweep Airplane Confi_Arations. NASA _4 X-142, 1959.

2. Alford, William J., Jr., Luoma, Arvo A. • and Henderson, William P. : Wind-Tunnel Studies at Subsonic and Transonic Speeds of a Multiple- Mission Variable-Wing-Sweep Airplane Configuration. NASA TM X-206, 19_9.

5. Spearman, M. Leroy, and Foster, Gerald V.: Stability and Control Char- acteristics at a Mach Number of 2.01 of a Variable-Wing-Sweep Con- figuration With0utboard Wing Panels Swept Back 7_ °. NASA _X-52, 1959.

4. Spearman, M. Leroy, and Foster, Gerald V.: Effects of Various Modifica- tions on the Supersonic Stability Characteristics of a Variable-Wing- Sweep Configuration at a Mach Number of 2.01. NASA_ X-260, 1960.

5- Foster, Gerald V.: Stability and Control Characteristics at M_ch Numbers of 2._0, 3.00, and 3.71 of a Variable-Wing-Sweep Configura- tion With Outboard Wing Panels Swept Back 7_ o. NASA_4 X-267, 1960.

6. Foster, Gerald V.: Effects of Spoiler-Slot-Deflector Control on the Aerodynamic Characteristics at a Mach Number of 2.01 of a Variable- Win_-Sweep Configuration With the Outer Wing Panels Swept Back 7_ o.

NASA_MX-273, 1960.

7. Spencer, Bernard, Jr.: Stability and Control Characteristics at Low Subsonic Speeds of an Airplane Configuration Having Two Types of Variable-Sweep Wings. NASA TMX-303, 1960.

8. Luoma, Arvo A. : Stability and Control Characteristics at Transonic Speeds of a Variable-Wing-Sweep Airplane Configuration With Wing Outboard Panels Swept 113.24 ° and 75 °. NASA _M X-342, 1960.

9. Foster, Gerald V., and Morris, Odell A.: Stability and Control Char- acteristics at a Mach Number of 1.97 of an Airplane Configtu_ation Having TwoTypes of Variable-SweepWings. NASA TMX-323, 1960.

i0. Whitcomb, Richard T.: A Study of the Zero-LiftDrag-Eise Character- istics of Wing-Body Combinations Near the Speed of Sound. NACA Rep. 1273, 1996. (Supersedes NACA_4 L92H08.)

Ii. Bielzt, Ralph P., Robins, A. Warner, ana Alford, William J., Jr. :

The Transonic Aerodynamic Characteristics of TwoVariable-Sweep

Airplsne Configurations Capable of Low-Level Supersonic Attack.

NASA TM X-30'_, 1960.

12. Spearman, M. Leroyj and Robinson_ Ross B. : Stability and Control Characteristics at a _ch Number of 2.01 of a Variable-Sweep Airplane Configuration Capable of Low-Level Supersonic Attack - Outer Wing Swept 75 °. NASA Ti,_ X-310_ 1960.

13. Rgoinson, Ross B., and Howard, Paul W. : Stsbility and Control Char- acteristics at a Mach Number of 1.41 of a Variable-Sweep Airplane Configuration Capable of Low-Level Supersonic Attack - Outer Wing Swept 79o and 108 °. NASA _4 X-52C, 1960.

14. Robinson, Ross B., and Spearman, M. Leroy: Stability and Control Characteristics at a Mach Number of 2.2 of a Variable-Sweep Air- plane Configuration Capsble of Low-Level Supersonic Attack - Outer Wing Swept TS, °. NASA _M X-330, 1960.

if) Z O)- ¢,\ ¢_..J • I I_,o S II II li i _ • ._lZ W., t-0 _z

SB

o ÷ ÷ _÷ Z ÷ Irr _" .... I-- w_

._N

u.l¢_ + ÷ ZG.

O _ I-- .J b. ¢_ mZ C._-I -r W )- _1 11_ _[ZUJ 2"/ >..

Z)

I--

el)

Z /

m

I--

/

=)

I.L

Z

| 2.6 / // / /

fJ)

//

//

Z

/ /

i/

G..

U.J

I

_J

I I:1:: U..

I

0 0 0 O0 0

o_o_

!

I I I _ -- ---_.

' • • " i "_ - • os_ oqp •.i '!" " '"" "'"

.. _ "i -:i aunlmum_....'i !.i

3o !

!

_0

Z

I--

-g

I

Z)

LI_

g

Z

V-

r_

W 3].

i

\ \ \ \ \ \ \ ,-g I

Z

\

I I,,--.

,,::C

r_ I I,.t.

Z

I I ¢0 I W if) _2 -.-I- ,-I o _J I I b_ !

I---' Z %,0

V

:D I.L.

Z I t"-- LL.

,° l,.,} "I- _L rr F- -r O_ '77_ - _'_ -_ ""'_:__";"_,'_ _" ! "" " HiP" il I ' " 0 • • , _1 I- m -0 m !

-COW

Z

cO r7

o_"

O.

w

1,1

_ObJ

W

Z

bL

I--

W

11.

I,

1,1

I ! I I I CY GO _ 0 X

EFFECT OF SWEEP ON SEA-LEVEL DRAG

M=I.2

DRAG

oM =0.9

0 _'J'_O"'°"-"_"_OM = 0.6

!

I I .... I I I

0 2O

40 60 80 I00 120

WING SWEEP, DEC

ee e_t • | I Q 3_ it 11

I

I

_o.

I

_{%1 I ! !

o

--: -- q _1 D6 W IJJ 6- Z Z O.

WN Iz.l Q. _s n," __Z ! v n .J _J :::) J 0 | --- N- _I W _o a m _.I a ::) -- (.)

z O. --J I I I Z J _o _" c:_e_ o_ I-- m

,el

!

..I r_ W_-_ fd) _OJ Z_ O_ W W_

I-

W W

<1

| @ @

E

_0 +1 ..J -I¢_ (:3 I!

W | | ,, I

J

r._ .,'t I I I t I t I I I !

!

!

!

I I I I " , I 0 ¢_1 "-

- Q

mQ / .I . . :. " :. ".'.

P.:".:" !i" " !.: III. _ONIC CRUISE AIRCRAFT Donald D. Baals, Cornelius Driver, and Owen G. Morris

N67-33073

Lsagley Research Center INTRODUCTION 'i%e a_/.a_ion world is on the threshold of sustained supersonic flight at a le'tel of efficiency approaching the best of our subsonic aircraft.

This potential has been recognized in the concept of the B-70. Now the supersonic transport is on the horizon.

The supersonic transport represents a great tec.hnical challenge.

Not only must this vehicle have the aerodynsmlc efficiency of the supL - sonic bomber, but it must also embrs_e: (a) the overriding element of passenger safety, (b) the problem of community acceptance (noise), and (c) econo_f of operation.

This paper will be devoted primaril_ to the aerodynsmic problems of large supersonic aircraft as related to performance. Smaller vehicles have been considered in part II of this compilation. The stability and control problems are of equal importance but are not discussed herein.

SfMBOLS A aspect ratio maximum cross-sectional area

%=

wette_ area incremental drag coefficient minimum drag coefficient wave-drag coefficient

%

mean skin-frictlon coeff.icient Cf llft ec efflele_t CL • • • @ • • • • •

..... : :: ..... : " i :'i

• _ , • • ee • • • tO .. ...... _...: t • • • • • • •e • .Q llft-curve slope

c_

chord c K proportionality factor for wave drag

Lp lift-drag coefficient

Z length L i M Mach number i m mass flow through slot i mass flow through a stream tube having the same cross-sectional

mo

area as the wing area fineness ratio n R Reynolds number S wing area SFC specific fuel consumption t thickness V volume W weight B=_-i S wedge angle #% angle of sweep DISCUSSION For long-range su_ersoalc aircraft the cruise efficiency is the primary design factor. From the Breguet range equation, which relates term is the lift-@rag ratio L/D. This ratio is _etermine_ by two basic factors - the minimum drag (ma_ wave amd friction drag at supersonic :, .,, . ...... . -.........

' ". ......... 4] ; ... .; ...

_ ! • te • # to speeds) and the drag due to lift. In figure 1 the relative drag breakdown for two representative bomber configurations in the 400,000 po_id-class is shown. For the subsonic configuration: the skin-friction drag and the drag due to llft are about equal, and the trim drag is small. For a Mach number 3 conflguration_ the total drag is about three times that of the subsonic ai1_lane_ Note that a new drag element, shock-wave drag, has been introduced. Even for she efficient configuration assumed, this ele- ment is _hout one-half the entire subsonic drag. Supersonic friction drag alone is equal to the entire subsonic drag.

Figure 2 points up the importance of aircraft geom.etrlc charac- teristics relative to the supersonic wave and friction drags. Shown here is the volutme coefficient V2/_/8 as a function of gross weight for various categories of aircraft. The volume coefficient is a meas- ure of aircraft fineness ratio and is an indication of wave drag. Note that the volume coefficient for a bomber configuration with its high payload density is less than that for a cargo aircraft which has a much lower payload de,.sity. Note also that, as the gross weight of the a_rcraft decreases, the vol_nme coefficient characteristically increases because the conflgaration is thickened to meet fuel and payload volume requirements. This condition places a wave-drag penalty on the small aircraft.

The wetted-area ratio characteristics shown on the upper half of figure 2 necessarily follow the same general pattern. Since the friction drag is a direct function of the wetted area, the parameter shown is a measure of the friction drag for a given skin-frictlon coefficient.

Figure 5 shows the variation of skln-friction coefficient for sdiabatic conditions as a function of Reynolds number and Mach number for both laminar and turbulent flow over an aerodynsmi_ally smooth sur- face. The turbulent curves have been experimentally verified for incom- pressible flow to a Reynolds number of 1,O00 × lO 6. At higher Mach numbers the skin-frlctlon coefficient is still in the process of evalua- tion. Note that there is little effect of Mach number on the laminar friction coefficient_ but there is a pronounced effect on the turbulent values. _,own on this figure are typical Reynolds number ranges for a Mach number 2 fighter, a Mach number 3 bomber-transport configuration wing at R _ 90 x 106 , and a fuselage at R _ 300 X 106 • Under laboratory conditions, the maximum Reynolds number for which laminar flow has been m_intained without some form of boundary-layer control is about 5 × lO 6 to l0 X lO 6. A maximum value of Reynolds number of 28 x lO 6 at a Mach number of 1.6 has been attained (ref. l) for a Q • . °.. .. : .....

e tee I • • _o e_ • • etl • • _-_ Qe cooled body of revolution, but the flow Was very sensitive to the slightest surface irregularities, even fingerprints. For a wing, the effects of leading-edge sweep also have been shown to be sdverse rela- tive to attainment of laminar flow.

There does appear to be a realm for application of boundary-ls_er control at subsonic speeds; however, supersonically, there appears to be little potential for attaining extensive l_mlnar runs - especially when the problem of traversing the pressure rise across a shock wave is considered.

L l The real problem of skin friction at supersonic speeds i primarily that of amtainlng the turbulent values for a smooth flat plate. Three- i dimensional roughness, surface waves, gaps, and so forth will tend to l increase the drag level as noted on the figure. The experimental data of reference 2 has shown the roughness effects to be less critical at the higher M_ch n_mbers. Current research is now leading to rational procedures for estimating the pressure and friction _rag for arbitrary types of roughness under turbulent conditions.

One interesting approach to reducing the turbulent skln-frlctlon values is illustrated in figure 4. This figure shows the unpublished results of tests obtained by John R. Sevier in the Langley _- by _-foot supersonic pressure tunnel on a two-dimensional airfoil at a Mach number of 2 wherein air was in_ected into the boundary Is_er to reduce the local- velocity gradients and therefore the skin friction. A substantial reduction in turbulent skln friction is noted, even if the initial drag penalty for the addition of the slots is considered.

Because of the momentum drag penalty, it is nct feasible to take air aboard to provide the Injection fl _; however, if low-energy air were already available, as from inlet bleed air, then this air might be efficiently utilized. Even bypass air me_ have appreciable energy; thus its momentum loss would have to be subtracted from the values shown here. The approximate amount of Inl_t bleed air available is noted in figure _ and this amount is shown to lead to a substantial reduction in skin friction. Whether this approach is practicable cannot be firmly stated at the moment, but further research and application studies are indicated. Certainly, the effect of Reynolds number on friction-drag reduction must be determined along with the effect of fall-chorcl injection.

Drag due to lift, as previously indicated, ma_ be the largest single element of the total drag _uring unaccelerated flight. As the cruise speed is increased into the supersonic speed ra_e_ the lifting m !

:. :...... ::..: .....

' • • : : 43 I ! I! • • ol • eo : , I ! i i i it i • e tii De Theories have been developed which indicate that large improvements are obtainable through the proper selection of planform and loading distribution.

Figure 5 is taken from reference 3 and summarizes the present state of theoretical P_owledge. :l"ne ordinate, the drag due to lift parameter _CD, is shown as a function of _ times aspect ratio for several wing.

plamforms. If this plot is considered to be for a fixed Mach number, L the _ term may be neglected, and all planforms will still have the i smme relative standings. These values are the minimum values of wave plus vortex drag as predicted by linear theory. The two-dimensional i i flat-plate value of 0.25 a.ld the minimum subsonic value of _ are shown for reference. The point to be made here is that large reductions in drag due to lift are indicated for the swept and the arrow wings, those for the arrow wings approaching the mini_ram subsonic value at low Mach numbers. Considerable experimental work has been performed in the Mmeh nu_er range from 2 to 3 for geometric aspect ratios of about 2 or for an adjlxsted _spect ratio of about 5 on this figure.

Figure 6 taken from an unpublished paper by C. E. Brown, F. E.

M_l_an, and E. B. Klunker summarizes recent work of the l_ley Research Center on a family of arrow wings. (See rel's. 4 mud 5-) The drag-rise factor -I _ACD is plotted against the parameter _ cot A, which specifies P eL,2 the position of the Mach line relative to the leading edge. For values less than i, the Mach line is ahead of the leading edge; at a value of i, it lies along the leuding edge.

The solid line (fig. 6) represents the theoretical drag-rise factor for an uncambered surface. _ne experimental agr_ememt (shown by the square symbols) is good for a wide range of values of _ cot A. Shown by the dashe_ line is the theoretical variation of the drag-r_se factor for a restricted camber loading. Here the best experimental values show only about one-half the anticipated theoretical gain. Subsequent analysls i_licated that thickness effects of the _- percent-thick biconvex airfoil could lea& to local supersonic flow, au_ shock-induced sept-" _tion would be _utieipated. Although oil-flow stud'_s &id not appear to indi- cate flow breakdown, there is hope that a revised wing employi_ _ouble- wedge sections in an attempt to eliminate supercritical flow might show substantial reductions in drag due to lift.

# # | # o _.

w | • $ $ @ 4 I # Another approach to improving the lift-drag ratio is the utilization of favorable interference. Although theoretical gains have been computed for mar_ unconventional approaches, most applications run into practical problems associated with real flows, poor off-desig_n chare2teristics, or large increases in wetted or base area.

Figure 7 shows the results of a _imple experiment at a Mach number of 3.11 to see whether the Lift-drag ratio of a 60 ° delta wing could be improved by the _Idition of a compression wedge to the under surface.

(See ref. 6. ) Pressures were integrated over the body and the lower L surface of the wing for a wide range of angle of attack, wedge angle, i and height-to-chord ratio. The resulting pressure drag is plotted against Lift. The base drag was not included in these integrations.

I i These results show no significant gain in llft-drag ratio, since most of the points fall above the experimental wing-alone curve. It should be noted, however, that, if a body or protuberance must be there in the first place, consideration of the interference flow fields can reduce the drag penalty. If base drag coul6 be eliminated by some form of base bleed or by filling the base with engine exhaust, then the added volume may actually have a zero drag penalty.

Up to this point, the elements of wave drag, skin friction, an_ drag due to lift, which comprise the llft-drag ratio, have been considered.

Now the levels of I/D currently attainable are considered.

Figure 8 shows the variation of llft-drag ratios with Mach number for a range of configurations. The level of the subsonic bomber a_l transport for fUll-scale flight conditions is shown on the left. For comparison, a marked decrease in the L/D level for current operational supersonlc aircraft is shown. This reduction is the direct result of the effect of the addition of wave drag and the increased drag due to lift characteristic of supersonic flight.

The experimental symbols shown (fig. 8) _re for wind-tunnel results at a Reynolds number of 4 × lO 6 for varlous research configurations representative of bomber types. (See refs. 7 to lO.) The exception is the swept-wing configuration _enoted by the di_or_ symbols, which has a volume coefficient representative of trausport-_rpe configurations.

The solid symbols are for complete configurations u_ler trimmed flight conditions. The open symbolm are incomplete configurations such as wing-body combinations. Note that there is a decided change in type of configuration being considere_ as a function of M. The highl_ swept, high-panel-aspect-ratio configuration tends to be optimum for a Mach number of 2 or less, whereas the low-aspect-ratio delta or trapezoidal planform predominates in the Mach number range from 2 to 4. Correction to i_all-scale Reyno_Is number of i00 × 106 has been mwle by assuming the boundary l_yers to be turbulent. A lift-drag ratio of about 8.7 is indicated at a Mach number of 3 for bomber-type configurations. A decre- ment in lift,rag ratio of from 0.5 to 1.O might be anticipated for con- version to transport configurations.

An estimate of future capability for bomber-type configurations has also been attempted as shown. The assumptions used are: (1) zero- lift wave-drag coefficient_ CDw = 0.0018. (2) friction coefficient lO j,cr- cent less than the turbulent value for a smooth flat ph:te, (5) wettcd- ZN3 D area-to-wing-area ratio of 2.8, and (4) drag-due-to-lift param,'tcr _CL _ from 0.14 at M = 1.5 to 0.18 at M = h.O. These estimates result in a level of potential lift-drag ratio of the order of lO at :: _:_ch numb_.r ML of 5. With such a gain, the level of crui_ _.ficiency _ would b_ SFC equal to or better than that obtained by the best of the current subsonic bomber and transports. With the other great _Ivantages of flight at a Mach number of 5, tbe leas-range subsonic aircraft would become techni- cally obsolete• Thus far, the on-design problems of supersonic flight, whic_ are primary considerations in the design of a long-range bomber, have been discussed. However, there are a whole host of off-deslgn problems applicable to the commercial supersonic transport which m_y dictate the design.

Part IV of this compilation shows that a typical commercial supersonic transport may consume onl_ one-half of its fuel under supersonic cruise conditions. The remainder is consumed in take-off, acceleration and climb, letdown, and in fuel reserves. These off-deslgn areas along wlhh other problems associated with passenger safety, Jet noise, ar_ the sonic boom must be solved without seriously compromising cruise performance.

Figure 9 illustrates one of %he off-deslgn areas which is critical for the supersonic transport - the landing and take-off problem. Plotted here is the variation of the velocity in knots with wing losding for a range of llft coefficients, that is, a simple plot of the llft-coefficlent equation. Shown for reference are the landing and take-off speeds for the present subsonic Jet transports - landing at about 125 knots and take-off in the range of 169 knots. For the wing loadings shown, these speeds represent a usable CL of approxlmstel_ 1.2 to l.&. Shown for comparison are some estimated values for a so-called "conventional" supersonic trans- port with an aspect ratio of about 2._. Even though the wing loadings are substantially reduced over those of the present subsoulc Jet, the io'_ aspect ratio and resulting low usable lift coefficient result in take-ofl speeds of the order of 200 knots with landing speeds of about 150 knots.

Also noted on this figure is an effectlve-aspect-ratlo scale. This is '_ ° o ° _ • , _ • • • • e Q • • • • • _ "" q_" • • • • • • @ • • :.! "! ! "i "'.: possible, since the effective aspect ratio defines tile lift-curve slope CL_ and therefore the usable CL for a given ground clearaz, ce angle. Flap effects are not included.

If present Jet transport speeds are to be considered a maximum from the landing and take-off speed standpoint, it is evident that a practi- cable supersonic transport configuration must attain higher trimmed £1f_ coefficients. This condition can only be provided through higher effective aspect ratio or a considerable advance in the trimmed flap effectiveness over that currently attainable. A variable-geometry wing has great potential in thls respect.

Another oft-design problem is that of transonic acceleration. The propulsion studies presented in part IV of this compilation show that the airfr_e-engine combination tends to be thrust marginal for the high-altitude transonic acceleratlorm dlc%atedby the sonic boom. (The boom problem will be discussed in part VIII of this compilation.)

One aeroclyuam_e element of this problem is the transonic wave drag.

Figure lO is a summary of the transonlc-wave-drag characteristics plotted against equivalent-body fineness ratio for a wide range of aircraft con- figurations, rsm_ing all the ws_ from the century series fighters at the low-fineness-ratio end to idealized research configuratior.s at the high- fineness-ratio end. Theoretlcally_ the wave drag is a functlon of the reciprocal of the fineness ratio squared. The ideal Sears-Haack len@th- volume body is shown to have a K-value of ll.1. Most of the airplane- type configurations settle out closer to a K-value of about lO. Detail refinement m_v be able to reduce this number, but it appears to represent a reasonable value for preliminary analysis.

With a minimum of assumptions as to aircraft size and geometry, it is possible to compute the variation of'the general drag characteristics with altitude for a range of key variables. Figure Ii shows the dra_ in pounds for a transport-type configuration at M _ 1. The plot on the left is shown for constant values of aspect ratio and fineness ratio for a range of values of maximum equlvalent-body cross-sectlonal area towing area. Note that the&rag reaches a minlmua at an altitude of about 59,000 feet and then increases rapldl_. At the higher altitudes • the drag due to lift predominates and wave and friction drags become secondary.

On the right-hand plot, the effective'aspect ratio (or more exactly, the drag due to lift) is varied. Aspect ratio is shown to be a power- ful parameter in reducing dra_ -. especi%_ll_ at the hlgher altitudes.

@ Dashed lin_s represent the general range of, thrust available for various Jet-propulsion systems which hav_ been sized for cruise at a i i • 3 ! a i , I D 4/ Msah rumber of 3- _.n accelerative force of 22,000 po'ind.(;, which wo,,*d produce an acceleration of 2 feet per second per second, ha_ beer, :n,b- %ranted from the engine thrust so that the values shown herr. indicate the net thrust available to overcome drag.

There is a critical transonic thrust-drag problem in t}.. alL.iLl:de rsr4ge of I_o)0oo to _0,000 feet. This altitude ms_f be the altitude dic.- tated by sonic-boom considerations. Th_ dropoff in engine thrust with altitude is so abrupt that the type of propulsion system and how it is matched to the airframe will be the primary factors determining .%c.,.,.l- eration altitude. At the _xtreme altitudes both th,: airi'ram,. ,_,_ro- dynamics and the propulsion system must be made optim_Am to attain the altitude levels desired.

Figure ]2 presents some cf the representative research coni'[t_- rations under study which illustrate certain aerodynamic approaches applicable to the supersonic transport problems. Configuration A is a low-ndnimum-drag, low-wetted-area conflg_ration made optimum for supersonic cruise. A double-bubble fuselage with the major axis in the horizontal plane improves the supersonic L/D. L_nding C L will be increased through the use of a variable-incldence wing. Co_,L'igura- tion B employs the lifting area rule. Its wing is ÷wisted and camb:.r,el.

Fuselage contouring and bodies located on the wing upper surL'a:,,, provid, favorable interference fields for the lifting condition. %_ne wing ms_v employ variable sweep to increase the landing CL. Configuration C employs outboard tails to improve the supersonic trim drag. _ne hori- zontal tails ridlr_ in the w_ng upwash improve the supersonic drag due to llft characteristics. Configuration D is a vari_le-sweep approach to the minimum drag supersonic configuration. A variable-d[hedral horizontal tall controls the transonic aerodynamlc-center shift and change in directional stability. Configuration E employs a blended w_ng-body approaah %rlth rapid thickness taper to reduce minim_n drag.

_%rlst _ camber are incorporated. Relatively thick outboar_ wing panels of variable sweep are provided. Co__figuratlon F considers a modification of the so-called "conventional" supe_rsonie transport to improve low- speed and transonic characteristics through variable geometry. The extensible tips shown provide about a 20-percent increase in area for subso_c flight. Alternate approaches employing variable sweep at the tip or substantial tip droop are a/so considered.

These configuration studies are merely illustrative of several aerodynsmic approaches under consideration fcr possible application to the supersonic transport. However, they do serve to illustrate the _erodynsmlc tools available to the designer to meet the extremely diffi- cult requirements of the supersonic transport.

il i ii:i i

SUMMARY There has been a gradual improvement in supersonic cruise effi- ciency to a level approaching that of the best of the present subsonic aircraft. There is still room fo, further gains in supersonic lift- drag ratio by fundsmental research on wave drag, skin friction, and dra_ due to llft. However, intensive research effort is required.

The most obvious problems are in the off-design areas of !_nding, take-off, t_-ansonic acceleration, and subsonic hold. T ese prob.emb L must be met without appreciable degradation of supersonic performance. l The most difficult problem, however, is the recognition that a problem exists in the first place. Consider that the total hours of supersonic i flight on all the B-98 airplanes number hut a few htundred, whereas the !

supersonic llfe of a successful transport m_y total as much as 25,000 hours. The technical problems inherent in such an advance are truly awe-lnsplring.

Although the problems discussed in this paper are broad and complex, there do not appear to be an_ obstacles which cannot be solved by a vigorous and effective research effort.

_9 o p pl_ • • ee ee • • REFF_R _;CES le Czarnecki, K. R., and Sinclair, Archibald R. : An Investigation of the Effects of Heat Transfer on Boundary-L%yer _rans1_ |on on a Parabolic Bod_ of Revolution (NACA RM-10) at a Mach number of i.61.

NACA Rep. 12_0, 1999. (Supersedes NACA TN's 5165 and 5166.)

2. Czarnecki, K. R. 2 Sevier, John R., Jr., and Carmel, Melvin M.: Effects of Fabricatlon-T_._ Roughness on Turbulent Skin Friction L st C.upersonlc Speeds. NACA TN 4299, 1958.

I 2 3. Beane, B. J.. Curves of Minimum Wave Plus Vortex Drag Coefficient for Several Wing Planforms. Rep. No. SM-22989, Douglas Aircraft i I Co., Inc., Nov. 1997.

4. _asson, Dennis _., and Wong, Norman D. : Aerodynsmi£ Characteristics at Mach Numbers From 2.29 to 4.65 of 80 ° Swept Arrow Wings With and Without Camber and Twist. NASA TM X-179, 1960.

9. Carlson, Harry W. : Aer dy_.amic Characteristics at Y_ch Mtmber 2.O_ of a Series of Highly Swept Arrow Wings Employing Various DegrePb of Twist and Cam r. NASA TM X-332, 1960.

6. Hasel, Lowell E. : An Experimental Pressure-Distzlbution Inve_ tigatlo.n of Interference Effects Produced at a Mach Number of _.I: by Wed_e- Shal_. Bo_ies Located Under a Triaag_lar Wing. Nt_ _ _M X-76, 1959.

7. Hill, William A., Jr. : Lift-Drag Ratios for Arrow _ings Alone and in Combination With a Body, Nacelles, and Vertical Tails a* Mach Number 3- NASA TM X-371, 1960.

8. Haliissy, Joseph M., Jr., e_i Hasson, Pm..mis F. : Aerodynamic Characteristics at Mach Numbers S.36 and ?..87 of an Airplane Configuration Raving a C_oered Arrow Wing Wit_ a 79 ° Swept Leading F_ge. NACA RM Lg_I, 19._8.

9. Church, James D., _es, William C., Jr., a_3 Sleeman, Wi]llam C., Jr.: Xnvestigation of Amrod_rnamlc CEaracteristics of an Airplane Configu- ration Having Tail Surfaces Outboard of the _g Tips at Finch _unbers of e.30, 2.97, and 3.51. NACA m_ L58CZ?, 1958.

IO. @_ant, Frederick C., :_ Sevier, John R., Jr.: Tra_oni- am_ Su_er- sonic _ind-Tunnel Tests of WlDg-B_ Combinations Designed for High Efficiemc_ at a ._mch Number of l._l. NASA TN D-h39, 1960.

m • • e • • .50 I • s > • • • oe iB eJs t CRUISE DRAG BREAKDOWN -TRIM L2%) 5O / 120°/,,I 4O - FRICTION (3401o I !

DRAG, LB_K) I% I- f-

_ TRIM

I01 I_ FRICTION

o

M=0.8 M=3 Figure i EFFECT OF AIRPLANE SIZE AWET -5 CDo = COW + Cf T °_-- I I 0 400 800 x 103 GROSSWEIGHT Figure 2 :'. :* ,. : :,. : ; ....

'.... ' _ii" iil ": : " ! | s s

SKIN-FRICTION CHARACTERISTICS

Figure .5 EFFECT OF BOUNDARY-LAYER INJECTION ON TURBULENT SKIN FRICTION M=2.0 R/FT = :3.5 x 106 1L SLOTS 4_,c -I00 MESH SCREEN I0 REI)UGTION IN Gf, PERCENT I0'_, ENGINE WEIGHT FLOW ! I FLOW COEFFICIENT, m mo Figure 4 | t ! • e=e • • • ee s e i 52 o • I I! e ee ! ! _ e • _ eee e _ i Let • DRAG DUE TO LIFT WING )ELLIPTICAL WING _'{LAT PLATE, DELTA WING ASPECT RATIO= oo t_ I v-= SWEPT WING, ro SONIC LEADING EDGE t-= ARROWWING, SONIC TRAILING EDGE I k-Tr ASPECT RATIO | - o J_ 0 5 I0 15 .8 xASPECT RATIO Figure THEORETICAL AND EXPERIMENTALVARIATION OF DRAG DUE TO LIFT THEORY FLAT PLATE WITHOUT SUCTION _.,'35/, ...... TWISTED ANDCAMBERED EXPERIMENT AIRFOIL - CIRCULAR a FLAT WING ARCc I=.025 o TWISTED ANDCAMBERED " 0 I "3ZI Oo "_ a i/"SONIC LEADING EDGE 0 A .e, = .8 !.0 ,8 COTA Fl_e6 "-' d 10,t O,,_,lb,l,qL.jl, ll_ _

<! _!:i :°

:• : :: I 4- INTERFERENCE LIFT AND DRAG .02

0%

CD PRESSURE

o Jr _-,,,.°'_°N_

@

0 CL PRESSURE Fi_e 7 AERODYN,_|C EFFICIENCY

%

L/D P T CORRF_.C_ED TO ,IIGI_ _,_ .............. _-.--;-.-.t ....... .j FULL 2 5 ' 0 I CRUISE MACH NUMBER F_,ur e 8 ....... ; '.. ,. ,::

i

Q _4 .: .o, ...... : CONFIGURATION EFFECTS ON LANDING AND TAKE-OFF SPEEDS . / c'C°_ L'- I I- I-

'°°t _

50 L _...t I I , i I , 0 25 50 75 tOO !25 WV, NG LOADING, W/S Figure 9 EFFECT OF FINENESS RATIO ON TRANSONIC WAVE DRAG

.o,[, \\

• °"!- %, o_°\ -- c_, .,-. "__

o_L,.\_: "_ " o',. _'_°_

.o, ___-"___ .;o

I -'_" -'--.- _ 15 Lt. t I ,, l I I I , I -= t I10 0 5 I0 15- FINENESS RATIO _lgu.t'e 10 I • _ e i t ¢ _ t ° • ), t w 4

" i

J Ip ; |'_ • ; ° e • I I ! | $1 Ill • • I OI, O t TRANSONIC DRAG CHARACTERISTICS (M_I) AWET W = 350,000; S= 4000 SQ FT = 35 7C_-x I03 • A = ,5 / AIT_X - .04 A-2 n=15

/

S n=15 \/ 50 AV_LAB_E \.a a

//

DRAG, LB 4O

/

-u.. n I I I I _ LI J eoxoo 3 !0 20 30 40 50 60XlO _ 0 I0 30 40 50 _2 ALTITUDE, FT Figure 13.

TRANSPORT RESEARCH CONFIGURATIONS ;'I Figure 12 IN_IqODUCTION In this part, the state of the art for several propulsion system components is reviewed, and some of the important factors which must be considered duri_ the design of a supersonic propulsion system are "pointed out. The discussion will be centered around the Mach number 3 cruise airplane of part III and the multimission airplane of part If.

SYMBO_

A_A_

ratio of free-stream tube area required by engine at any Mach

AM= 3

number to corresponding area _t design point of M = 3 h altitude M Mach number mass-flow ratio, boundary-la_er bleed P static pressure total pressure Pt

p_

Inlet control parameter Pt angle of attack 5_ " " : : ." " " " " " e • • e °o_ e|| ! I no ee DISCUSSION The pressure recovery characteristics of supersonic inlets have an important effect on propulsion-system performance because for every per- ¢_ntage of loss in recovery there is a greater percentage of loss in net thrust, q%e recoveries which have been obtained a_ zero angle of attack from several types of supersonic in!ets 8me shown in figure i. The inlets include two-dimensional 8md long axisymmetric types with large amounts cf internal compression and a short axisymmetric type with equal amounts of external and internal compression. The off-design spillage drag of L the long axisymmetric inlet is considerably less th_n that of the short i axisymmetric inlet. These inlets were de_,i_ed for M = 5 and all 2 have variable geometry provisions - either as ,_riable ramps or as i translating center bodies. The inlets, as sketched, have equal capture _reas, and the sketches therefore indicate the relative lengths of the supersonic diffusers. The two-dimensional inlet has the highest recovery, varying from0.89 at M = _ to 0.94 at M = 2. These values are qu_e high and further significant increases of recovery will not be easily attainable with any inlet design. There is no fundamental reason why the recovery of the long axisymmetrlc inlet should be less than tl_t of the two-dlmensional design. The difference shown in figure 1 merely indicates that additional development time must be spent on the lor_ axlsymmetric inlet. All of the inlets incorporated some form of boundary-layer bleed. The amount of bleed at M = _ was from 7 to 12 percent of the inlet flow. Additional boundary-l_yer removal tP_ou_ the existing bleed system did not significantly improve the performance of the long axisymmetric design at M = 9- Because of the drag penalty associated with the bleed air, the optimum recovery - that is, th_ recovery at which the maximum value of thrust minus bleed drag is obtained - may be less than the maximum recovery. Our present knowledge of the behavior of the turbulent boundary layer in the presence of adverse pressure gradients is not sufficient to be able to predict exactly the effect whlchbounda_-layer bleed will have on the increase of pressure recovery. Therefore, a detailed experimental tailoring of the bleed system willbe necessary for each inlet to develop the optimum design.

These i,_lets, in addition to baying a hlgh pressure recovery, must be able to supply the required englne airflow. The airflow character- istics of several turbojet and turbofan e_ines are presented in figure 2.

These engine airflows, as presented, are a function of prezsure recovery a_d .are based on the recoveries of the two-dlmensio_al inlet. The reco,ertes have been extrapolated to a value of 0.96 at M = 1.0 (fig. 1).

In figure 2 the airflow variation is expressed as the ratio of the free-strea_ tube area required by the e_lne at any Math number to the corresponding area at the design point of M - 3 and is plotted as a .o.

, . • .. . ..

. , , . " 5"3 e t function of _ch number. The shaded area indicates the envelope ot the airflow characteristics of a series of turbojet engineJ and the cross- hatched area reprezents tne corresponding envelope for turbofan ermine.',.

The differences of the airflow characteri:;t[cj of the turbojet engines are small throughout the Much n,snber rang.z, but a large difference exists between the t,Arbofan engines. The airflow characteristics of the two- dimensional and short axisylmnetric inlets shown in figure ! have been included here to indicate that the inlet and engine airflow charac!cri_- tics have similar trends with Much numbcr. Thcsc inlct _irflows can, of course, be controlled to vo.rylng degrees during the inlet design to match the airflow of a specific engine. It should be mentioned that the comparison between the engine and short axisymmetric inlet airflows is not strictly correct because of the lower recoverj of this inlet.

Accounting for this effect would increase slightly the stream _ube area ratios of the engine_ at off-design Much numbers.

All of the engines require a significantly smaller stream tube of air at low Much numbers than at the design speed of M = _. Thiz excess air must be diverted from the engine, either by spillage ahead of the inlet or by bypass ducts located_hind the inlet. Regardless of how the air is diverted, a drag penalty will be incurred.

The magnitude of this matching drag penalty is now exsmdned. The discussion will be based on a turbofan airflow characteristic defined by the top of the cross-hatched area in figure 2. At any off-design Much number the distance between the top of the cross-hatched area a1_ the design value of 1 represents the amount of air which must be spilled or bypassed. At M = 1.2 this amounts to 41 percent of the design stream- tube area.

The matchlng drags created by four typzcalmethods of di_rtlng the excess air are presented In figure 3. These drags .have been divided by the thrust of a typical turbofan engine and are p_ented as a function of Much number. The sketches on the left side of the figure depict two typical ways of spilling the air ahead of the inlet - behind the _hock wave of a 5 ° translating wedge and by means of _he flow field of a i0 ° half-angle translating cone. On anlnlet these shapes would be the initial compression surfaces. The excess air may also be bypassed from the subsonic diffuser by means of sonic nozzles which are parallel to the airplane axis or are inclined at some angle, such as lO ° . It has been aJsumed in these drag calculations that the total pressure recovery at the exit of the sonic nozzles was equal to the inlet recovery.

The. spillage drag is generally largest at transonic speeds, end it is also most critical at these speeds because the value of the. accelera- ---_ thumastmznus drag is 5mmll, aS wx_ be s_vn subsequently.

6o i i i | i ii ! j , ii The least d_'ag at transonic speeds is incurred by the use of the sonic nozzle, exii.ing axially, and the drag penalty is about l! percent of the engine thrust. A word of caution is in order here. An increase in frontal area may be necessary to bypass the large amount of excess air at transonic speeds and exhaust it in an axial direction. The drag penalty associated with this increase in frontal area is not accounted for in this comparison. The highest drag is created by the _o wedge and is about i0 percent of the engine thrust. Spillage ahead of the inlet can be accomplished more efficiently with the i0 ° cone than with the L 9 ° wedge. It is also apparent that bypassing the air through a sonic nozzle inclined to the axis results in an appreciable rise in the bypass 2 ' drag. At Much nurJbers from 2 to 3 the matching drag is smaller when the air is spilled ahead of the inlet instead of being bypassed from the subsonic diffuser by a sonic nozzle. Use of a supersonic bypass nozzle would result in bypass dlags which were less than those of the i0 ° cone at supersonic speeds but would not result in drags which were lower than those of the sonic nozzle at transonic speeds. It is of interest to note that the matching drag of the short 8atlsy_mmtric inlet is quite high.

For example, at M = 1.6 the calculated value would be about 0.094 for the airflow spillage considered in figure 3. The magnitude of these matching drags in terms of engine thrust is a function not or_ly of the airflow being diverted but also of the engine thrust per pound of air- flow. For an afterburning turbo.* ,, the percentage loss of thrust per pound of air diverted would be about half the corresponding value for turbofan engines.

In figure 4 the percent of change in the zero-angle-of-attack recovery h_s been plotted as a function of angle of attack. 1_ese data indicate that at a Mach number of 3, the pressure recovery may be reduced by 6 to 18 percent by _n angle of attack of only 4°. These percentage reductions become smaller as the Math number is reduced.

With a two-dimensional inlet, the angle-of-attack or angle-of-yaw effects m_y be reduced by use of a horizontal or vertical compression surface, respectively. Such an arrangement is no_ possible with a three- dimensional design. It appears highly desirable that the supersonic inlet, regardless of its type, should be shielded from flow-angularity effects as much as possible by placing the inlet in a flow field gener- ated by the airplane wing or fuselage. Such _n arrangement may also reduce the inlet size, reduce the amount of supersonic compression which the inlet must accomplish, and simplify the inlet-control system.

The controls for supersonic inlets form a very vital part of the propulsion system. Generally, both contraction and bypass controls are required. The functions of these controls are to obtain high-pressure recovery and to prevent shock regurgitation which results in large decreases in pressure recovery, inlet buzz, and the attending engine surge. Shock regurgitati, I should be prevented for a number of reasons.

v f d ...... . ..... :,,.

• . .._ • . : :'" :', : : • " : :" : :" " ' 61 60 ea oe • • • o,9 " A combination of the discontinuous drop in recovery from perP_ps 0.09 to O.D9 and the buzz tPmt usually follows shock regurgitation will cause the engine to surge at Mach numbers near 3. This means that most if not all of the thrust of the unstarted system wlll be lost and inlet drag incurred until the inlet can be restarted and the engine returned to normal operation, which may require a complete shut down and restart.

In addition to the destabilizing force associated with this thrust loss, the inlet will spill large amounts of flow which could affect operation of nearby inlets or cause a high pressure region under a nearby win 6 or on the body. Further, a structur_1 problem is caused by the engine surge. Duet static pressures in excess of free-stream total pressl_e have been measured during engine surge in a turbojet-engine-inlet con- figuration. These pressures are much higher than normally exist in the inlet and require a considerable increase in the structural weight over that required for normal operation.

The contraction control is required to vary the inlet throat for changes in airplane Maeh number and angle of attack or yaw. Because the effect of Math number and angle of attack on optimum contraction differs for the various inlet types and with position on the airplane_ a particular control will probably have to be developed for each specific application. Attempts to measure a throat supersonic Mach number for use as a control parameter have been frustrated by the terminal shock affecting the static pressure throughout the throat region. Therefore, since no contraction-control parameters with general application have been observed, it may be necessary to schedule the inlet contraction as a function of one or more parameters that are indicative of airplane M_ch number and angle of attack or yaw. The problem of sensing flow angle can be simplified by sheltering the inlet under a wing or near the body, thereby restricting major flow-angle changes to either yaw and cross flow er angle of attack.

The contraction control is required to be fast acting to prevent shock regurgitation during airplane maneuvers and gusts. These extermal disturbances cause temporary errors between the desired and actual con- traction; therefore, the desired contraction must be less than the opti- mum value by a margin equal to the maximum expected error to prevent shock regurgitation. Since these errors can be reduced by increasing the complexity of the control, a compromise must be made between the control complexity and the performance margin required for stable inlet operation.

The bypass control positions the normal shock near the inlet throat to obtain high recovery and at the same time matches the inlet and engine airflows by spillage ahead of the inlet or by a bypass in the subsonic diffuser. For this type of control, the problem is to position the terminal shock as far upstream in the throat as possible, without allowing airplane maneuvers or engine transients to force it forward of the throat. The entire bypass control loop, which includes the sensor, • • _o oeO 62 "" " : '_ : oo cot , IJu oo eo • oo control, bypass actuators, and the response of throat conditions to bypass movement, must be anal2zed to determine the decrement in recovery required for stable inlet operation. The response of throat conditions to b_ass movement, or duct dynamics, has been measured for a number of inlet-cold-plpe and inlet-ermine configurations. Good agreement has been obtained be%ween these measured duct dynamics and a simple prediction based on a dead time equal to the acoustic travel time from the bypass to the throat, in series with a first order system based on the ability of the diffuser to store mass. In order to keep the recovery decrement required for a stable, controlled inlet operationsmall, the throat flow L conditions must respond quickly to bypass movement. In order to obtain i this fast response, the prediction shows that the bypass must "De placed near the throat to minimize the dead time and the diffuser volume must be i small to reduce its storage capacity.

In order to illustrate possible bypass control sigr_ls, static pres- m_e distributionz on the centerbody of an axially sy_netric inlet with flush slot bleed are shown in figure 5. In the plot on the left side of the figure for a Maeh number of 2.88, the shock moves up to and is com- pressed on the flush slots as recovery is increased. Just as recovery is increased from 0.851 to the peak value of 0.8_, the shock begins to mo_ ahead of the bleed and appears to furnish a control signal with large gaiu. However, at a Mach number of 2.48, the terminal shock does not travel ahead of the slots as peak recovery is reached so that no static pressure rise as large as that across the terminal shock is avail- able for control purposes. Therefore, when the problems of off-deslgn operation and the need for a supercritlcal margin to obtain stable inlet operation are considered, it appears that in this inlet it is not possible to place a sensor where the terminal shock will consistently pass it at conditions near peak recovery. This was also found to be true for the two-dimensienal inlet (fig. l) with porous bleed because althou@h the axial position of the geometric throat was constant, the peak recovery shock position was found to be a function of contraction, Mach number, and angle of attack. The throat bleed _nd boundary layer, thus, seem to distort the throat flow so that direct shock-posltlon sensing does not seem feasible. However, for this axially symmetric inlet, the static pressure downstream of the slots appears to vary continuously as recovery is chan@ed at both Mach numbers of 2.88 and 2.48. A control, which sensed a constant ratio between the statlc-pressure sensor Just downstream of the slot and a throat total pressure, set recovery within 2 percent of its peak value frc_ Mach numbers of 2.0 to 2.88 at zero angle of attack.

A constant ratio of throat-exit static pressure to throat total pressure was also found to be a very satisfactory control parameter for the two- dimensional inlet Just mentioned. The parameter is, therefore, believed to be of general use and is equivalent to setting a constant throat-exlt Mach number. Summarizing the inlet control situation, it appears that considerable development effort will be required for each different inlet installation.

b-- °._ ee pro e • ! e! ) _ ! ij! ! i I ! i _- L ,, 6_ • e s | Regarding the exhaust system, reasonable performance can be obtained from a conventional sonic exhaust nozzle at subsonic speeds. For the engine which must operate at supersonic speeds, however, varlable-geometry ejector nozzles are required. The performance of such a nozzle is shown in figure 6. Here the ratio of the net thrust minus boattall drag divided by the ideal net thrust is plotted as a function of Mach number. The per- formance is shown for a design which has primary and secondary nozzles of variable area and a variable external shape. The net thr.:st ratio reaches a minimum value of 0.82 at transonic speeds and gradually increases to 0.91 at the design Mach number of 3. Inasmuch as the ideal variable- geometry ejector has a thrust ratio of about 0.97, it is believed that considerable improvements in ejector performance, particularly at tran- sonic speeds, may be realized with further research.

_l of the propulsion-system components which have been discussed A_sult in losses in the thrust which is available from an engine. The magnitude Of these losses is shown in figure 7 in which the ratio of net thrust to ideal net thrust is plottad as a function of Mach number.

This breakdown includes the losses due to the two-dimensional inlet pressure recovery, inlet control margin, boundary-layer b _eed, spillage, and the variable-geometry ejector just discussed. _le largest individual losses are due to the inlet recovery and variable-geometry ejector. The losses due to control margin, boundary-layer bleed, and spillsge drag are each small, but t_e sum is significant, and such losses are inherent in any propulsion system. As a result of all these losses the available thlust is from 70 to 75 percent of the ideal thrust. Increases in the available thrust ratio above these values will depend principally upon future improvements in inlet and exit performance.

Turning to the engines themselves, both tur_oJet and turbofan engines are being considered for use in the multimission and supersonic cruise airplanes. As is known, the basic gas generator for these two types of engines is essentially the same. I the fau eL.;ine, however, extra power is extracted from the main gas stream by a t_rbine to compress additional air, which does not pass through the gas generator, so that for a given gas generator the total airflow of the turbofan is greater than that of the turbojet.

When the power plants are considered, the : hrust characteristics of the engine must be compared with the drag characteristics of the airplane.

This comparison has been made in figure 8 for the supersonic transport.

(The comparison would be somewhat different for the supersonic bomber. ) The airplane was assumed to accelerate to M = 0.9 at low altitude, climb to 40,000 feet at M = 0.9, and then climb gradually while accele- rating with a u_mente_d pow__er to M = 3,0.

The line labeled drag represents the level of drag encountered during the climb and acceleration, and the symbols represent the loiter, rA" ..:":: 64 : '. ".

I¢ |l subsonic-cruise, and supersonic-cruise drag values. The two shaded regions represent the levels of thrust obtainable by typical turbojet and turbofan engines matched to this airplane with a nonaugmented take- off thrust-to-weight ratio of O. 31. The band on the left is for unaug- mented operation; the other band is for full augmentation. The downward Jog between the two bands is, of cours% associated with the climb to altitude.

The most significant thing shown in figure 8 is the fact that the thrust-minus-drag margin is a minimum at transonic speeds and is much L smaller than the margin which exists at either subsonic or supersonic i speeds. This characteristic may well determine the required engine size.

Other problem areas are take-off noise (which is rather high for the i turbojet even "_ithout augmentation), the fact that sea-level loiter and

subsonic cruise are performed at extremely low percentages of the avail- 3

able unaugmented thrust, and the necessity for the attainment of low specific fuel consumption at supersonic cruise conditions.

A large percentage of the fuel usage of the supersonic transport occurs in off-design flight. This is illustrated in fig_ 9 where the fuel rate for such an airplane is plotted against flight time. The mission segments considered are climb and acceleration, cruise, let- down, loiter, and reserves. Only about half of the fuel is used during design-point operation. This fact stresses the great need for obtaining efficient propulsion system performance in off-design operation as well as at the design point.

The same situation exists in connection with the multimission air- plane. In figure lO the percentage of fuel used by such an airplane during the various phases of three important missions is shown. The three missions are: a M = 1.2 sea-level-dash mission, a supersonic- cruise mission, and a subsonic-ferry mission. In each case the mission is broken down into take-off, climb_ and acceleration, the dash or cruise part of the mission, and the loiter and landing. It is apparent that the major part of the fuel usage for the three missions occurs at quite different operating conditlons: M = 1.2 at sea level, M = 2.2 at altitude, and M --0.85 at altitude. Here again an extremely versatile engi_ is required for this airplane. In other words, _he engine should have a low specific fuel consumption for a wide range of operating conditions.

The basic characteristics of turbojet and turbofan engines are c_- pared in a qualitative sense in figure ll, where specific fuel consump- tion is plotted against thrust for several operating conditions. The turbojet engine data selected are representative of current M = 3.0 designs, whereas the turbofan data have been obtained frQm industry and NASA estimations of the probable characteristics of such engines.

two curves on the left are for M = 0.9 flight at 35,000 feet with no augmentation, and the other t:_o curves are for M = _.0 flight at 69,000 feet with augmentation. The two engines have been sized arbi- trarily to give the same thrust at take-off conditions.

At subsonic speeds it can be seen that the turbofan potentially offers lower specific fuel consumption and thus more efficient operation over a much broader range of thrust than the turboje t. This cl_rac- teristic, which is obtained with little if any penalty in specific fuel consumption in the supersonic-cruise condition, obviously is of great benefit at off-design operation such as subsonic cruise and loiter. In L addition, because of the greatly increased airflow the turbofan poten- tially offers advantages of lower noise during take-off, greater thrust au_men_tion at transonic speeds, and lower operating temperatures during both acceleration and supersonic cruic One disadvantage of the turbofan is that it might require a greater frontal area for a given thrust and will have greater air inlet and ducting weights. Obviously, therefore, there is the problem of trade-offs - that is, the optimum engine for a given airplane or mission can be determined only by a step-by-step consideration of all the factors involved. Nevertheless, the potential advantages of the turbo- fan appear great enough fc both the supersonic and the multimission airplanes that its develo_,.ent should be pursued vigorously.

In summary., the state of the art with regard to air-breathing pro- pulsion systems for supersonic airplanes may be expressed as follows.

Sufficient research and development has been conducted in the fields of ai_. • inlets and jet exits to enable a reasonably high level of on-design performance to be obtained. A great deal of detailed tailoring will be required, however, to match components so that optimum performance will be obtained over a wide range of operating conditions, and satisfactory control arrangements worked out. In connectlor -ith the engines them- selves, the turbofan appears to offer a number o, significant advan- tages which appear to make its future development highly deslrable.

,:,,: : -,, : ..... : ..... : • • • eOo ee o6 ....... :. -°11 66 ll J • • II I1| I I I 11 l qm SUPERSONIC INLET PRESSURE RECOVERY a:O ° 1.0 .9 TOTAL .8 PRESSURE I RECOVERY .7 I-., p., .6 LONG AXISYMMETRIC

.SL

!_[ i I 0 i 2 3 M Figure i TYPICAL AIRFLOW VARIATION STREAM-TUBE AREA \, \ "_ TWO-DIMENSIONAL ._ " INLEt\

s

•

• INLET

.3LI

i I I , , I 1.5 2.0 2.5 3.0 O I.O M Flsure 2 _ J w J IL I| i It | _ ot Do t, e • OlD@ _a" MATCHING DRAG FOR TURBOFAN ENGINE .12

|

.I0 _,_ILLAGE DRAG _ BYPASS DRAG-SONIC NOZZLES 5° TRANSLATINGWEDGE &t_TCH I NG DRAG ENGINE THRUST IO° _ NG .04 .02 I 2 M Figure 3 EFFECT OF ANGLE OF ATTACK ON PRESSURE RECOVERY M=3 PERCENT CHANGE 0 = 0 ° PRESSURE RECOVERY +10 LONG AXISYMMETRIC SHORT AXISYMMETRIC I0 0 I 2 3 4 5 6 7 8 9 Figure 4 VARIATION IN CENTERBODY PRESSURE DISTRIBUTION WITH RECOVERY P CONTROL PARAMETER- p_ CENTERBODY !

/.-, STATIC PRESSURE _N M = 2.48 !

Figure 5 VARIABLE EJECTOR PERFORMANCE NET _RUST-DR_ IDEAL NET THRUST .S _'Xt-,,.. ,_ M < 1.00 M_?.O I. I I • 0 IO EO M Figure 6 I 3D

....... : !! :."

THRUST LOSS BREAKDOWN NET THRUST IDEAL NET THRUST !.0 - ,__-INLET PRESSURE _ RECOVERY LOSS • 9 J_ - " __-- " " -__ NLETCONTROL MARGIN LOSS , r "---BOUNDARY-LAYER BLEED LOSS '-SPILLAGE DRAG LOSS °9 • • ,_ (10 ° CONE) VARIABLE- GEOMETRY EJECTOR LOSS ,-I oJ ,--i I "_" NET THRUST M Figure 7 SUPERSONIC TRANSPORT AIRPLANE L WITH -_ RELATIVE THRUST AND DRAG

/-°-°

SUBSONIC Gnu I SE

su_.o,,c /

. "CRUISE _ • z ! !

I 2 3 O Fi_re 8 r- ..... :.,,:..

• • q 7O • • Qe e el e t • • • e , t e • • ees ! I o s ] / SUPERSONIC TRANSPORT AIRPLANE FUEL USAGE FUEL FLOW RATE

_J

F CLIMB AND ACCELERATION

CRU,SE RrRES s

I _I I" I

TIME F1gu.z'e 9 MULTI MISSION AIRPLANE IOO • M, 1.2 SE.A LEVEL _LSH k"q SUPERSONIC CRUtSE _1 FERRY FUEL USED % r]BIB ........ n.

O L TAKEOFF, M • I.E) M-220 M,0.85 M,O.50 AT CLIMB, AND AT AT AT SEA LEVEL ACCELERATION _ LEVEL ,4_ITUDE 8d.TII'UD[ I.ANOIN6 I I i i i _ .

ENGINE CHARACTERISTICS TURBOFAN .... TURBOJET I" M=5.O; SUPERSONIC i _" _ h=65,000 FT;, CRUISE_ION SPECIFIC FUEL C/ONSUMPTION SUBSONIC CRUISE i RELATIVE THRUST Figure ii _.I_C_OlNG PAGE BLANK NOT FI_,:._C-.

• 3 • v. _E x-15 __.ICHT RESEARCH _O_i_ '_ RE_TTON TO' THE DEVELO_ OF ADVANCED MILITARY AIRCRAFT By Jack Fischel Flight Research Center

N67-33075

INTRODUCTION H The design, construction, and development of advanced military air- craft undoubtedly will involve the utilization of many relatively new concepts in several different, but related, areas. In many cases, these new concepts will require new flight techniques or involve development problems inherent in the use of relatively unproven methods, materials, structures, systems, and configurations. Moreover, flight verification will be required of aircraft aerodynamics and flight behavior because of the usual uncertainties associated with predicted data. Solutions to these problems obviously would require flight testing involving exten- sive time and effort before the aircraft could become operationally acceptable.

Upon the inception or the X-19 research airplane project, many new and far-reaching techniques and principles were studied and applied in the aircraft design and are currently being demonstrated and investi- gated. Because similar concepts are likely to be used in advanced mili- tary aircraft, the X-15 flight program will provide significant informa- tion, over a broad flight environment, pertinent to the development of these vehicles. Tais paper discusses the research objectives of the X-15 flight program, some of the flight aerod_amlc characteristics currently being obtained, some development problems encountered, and the experience obtained with the advanced systems investigated.

SYMBOLS 4b_ag coefficient

%

lift coefficient C L C_ rolling-moment coefflc lent

i

Z._, 74 : "" ........ :... :..

!" .:: .!!: .." ! • ....

.: ,.: : ' . '. ......... : ",: C_5a = _Sq pitching-moment coefficient C m H i normal-force coefficient CN Cn yawiug-moment coefficient Cr_v = _v M Mach n_mber q dynamic Im'essure em.Ele of attack angle of sideslip aileron deflection 5 a horizontal-tail deflection 5 v vertlcal-tall deflection q ' ,i" ! i' !::. !.: OUTLINE OF FLIGHT PROGRAM Research Objectives In order to provide some understanding of the contributions forth- coming from the X-15 _li_t program, a listing of the flight research objectives is presented as follows: (1) aerod3caamic and structural heating, (2) aerodynamic loads and struct'_-al research, (3) aerodynamic deriv_,tives, (4) flight control, (9) lift and drag characteristics, (6) recovery and land_ng, (7) aeromedical studies, and (8) operational evaluation. These ite_ are discussed individually in the following paragraphs.

In regard to the i_em (I), aerodynamic and structural heating_ one of the primary objectives of the X-19 flight test program is to make heat-transfer studies on a full-scale flight vehicle in the true envi- ronment. In order to accomplish this objective, temperature gradients in the structure wil] be obtained as well as the isolated skin tempera- tures from which heat-transfer coefficients could be determined. During the flight tests, primary emphasis will be placed on studies of the aerodynamic heat transfer to the vehicle, particularly in the areas where interfering flows are experienced and where available analytical methods might be expected to give less satisfactory predictions. The flight results will be nondimensionalized wherever possible so as to be of most general applicability. Detailed studies of boundary layer and local flow conditions will be made in selected areas on the airplane in order to accomplish this nondimensionalizing.

In the area of aerodynamic loads and structural res_rch, aero- dynamic and structural loads data are being obtained on the wing, con- trol surfaces, and various structural components of the X-19 airplane during flights and also during landings. This information is of interest to structural designers and aerodynamicists in p:-oving the integrity of structures and predicted loads characteristics and the efficiency of control surfaces under varying environments.

Determination of aerodynamic derivatives in flight over the oper- ating envelope is of obvious significance for verification of wind- tunnel and estimated derivatives and for flight-planning purposes.

The flight control research areas are manifold and include inves- tigation of control problems in supersonic and _ypersonic flight; con- trol problems in traJecto1_ flight, including exit, control at low dycamicpressure and nee_ zero g, an_reentry; advanced flight control systems, and control in the Dyna-Soar research areas. Among the signif- icant objectives included are: determination of the adequacy of the 0/splay, of the console stick, and of the stability augmentation system; i l|e • I. • • • 76 ....

...... :: the use of reaction controls; simulation requirements; handling qualities in the hypersonic regime, in a hlgh-dynamic-pressure high-temperature environment, and in a low-dynamic-pressure hlgh-angle-of-attack regime; and a study of a type Of adaptive control system for effective operation in an extreme control environment.

Research on lift and drag characteristics involvesdeterminatlon of Ecale effects on lift and drag. These effects, of course_ aid in the verification or interpretation of wlnd-tunnel results and extend the knowledge of Reynolds number effects. Also, base-drag character- istics are being determined.

The recovery and landing research performed is applicable not merely to the safe landing of the X-15 but to the proper determination of suitable recovery techniques for use on vehicles having low lift- drag ratios, and, certainly, is applicable to the Dyma-Soar.

Aeromedlcal studi_s involve determination of physiological aspects in a varied g environment or 4_ _-_tical regimes. Another impor- tsnt aspect is the flight evaluation of a pressure suit made to with- stand the range in temperature and pressure and the blast effects anticipated for normal and emergency operation.

Operational evaluation is applicable to the various systems, mate- rials, and structure utilized in the aircraft and will provide informa- tlonpertinent to these items in a varied flight enviromment.

Performance Envelope The extent of the flight regimes available to the X-15 is shown in figure 1 in terms of altitude audMach number. The flight envelope obtainable with the XIRll interim engines extends to a peak altitude of 135,000feet and a Mach number of 3.4. The design altitude w_h the XLR99 final engiue J s 250,O00feet. The design speed is a Mach number of about 6.5. Higher altitudes at lower airspeeds, as shown by the dashed curve, can be achieved in acute semiballistic or ballistic flight trajectories. For flights to extremely high altitudes, however, recovery is uncertain because of the reentry problem.

Until now, the X-15 has achieved a peak Mach number of 3.2 and a peak altitude of 107,000 feet by using the interim rocket engines.

Further expansion an_ exploration of the fligh_ envelope available with the XIRII engines will continue in the next few months, with special emphasis on several of the research areas discussed, such as heating, rocket engine is currently be_g installed in one of the X-15 airplanes an_ will be used in the near future to expam_ the flight envelope to the a_ eeJ_ • _ Ds iIJ • • e| ) • O • I ; s I O0 $el I; airplane design limits. Exploration of the research areas discussed will be _ursued simultaneously within this flight envelope after all the airplanes have had the larger engine installed.

RESULTS OBTAINED TO DATE As is well known, a substantial amount of theoretical and laboratory research, as well as simulation, was performed in support of the X-15 project. Therefore, verification is required of these predicted charac- teristics under full-scale flight conditions. The operational experience with systems, techniques, and materials, also requires evaluation.

Stability Derivatives and Flight Control The current status of the flight evaluation of some of the principal stability derivatives is shown in figure 2 as a function of _ch number for a limited angle-of-attack range (4° < 6 < I0°). Shown are the lift- curve slope and the longitudinal-, directional-, and lateral-stability parameters. For comparison, wind-tunnel data for a similar angle-of- attack range are also shown, by the faired curves, and indicate fairly good agreement. The pitch, yaw, and roll control derivatives evaluated in flight are compared with predicted results in figure 3. Good agree- ment is also indicated. Although good agreement between flight and wind- tunnel derivatives is apparent in the range below a Mach number of approximately 3, particular interest is centered in the flight-derivative evaluation at Mach numbers in excess of 5, where no previous flight eval- uations have been made. Verification of wind-tunnel derivatives at these higher speeds will provide information pertinent to the development of other advanced vehicles.

It is of some significance to note that various flight motions have been reasonably well predicted in simulator studies thus far by using wind-tunnel derivatives. This agreement has provided a degree of assur- ance in expanding the flight envelope. Although the stability augmenta- tion system has been used inmost of the flight studies performed, fur- ther research is planned with and without the use of various damper modes to evaluate the augmentation system and to determine amy control limita- tions resulting during normal operation or from damper-out conditions.

In specific flight regimes, such as at high d_namic pressua.e or during reentry_ control limitations maybe critical; therefore, the X-I_ sho1_Id provide information applicable to minimum control requirements of other advanced vehicles.

, j o I 9 • e e •

78 :" "

::. . ". ". ".. ".•: ..... " 'i'!

oo ooe • e • • • oe4_ 06 4o • Lift-Drag Characteristics A comparison of flight and wind-tunnel drag polars at three speeds is shown in figure J4. Presented are the variations of lift coefficient with drag coefficient for Mach numbers of 0.9, 1.1, and 2.0. Agreement is reasonably good at the subsonic and higher supersonic speeds, but the fllght-determined drag coefficient is somewhat higher at the lower supersonic speed. Brief base-pressure data indicate that this disagree_ ment in the low supersonic range is due largely to a discrepancy between the base-drag measurements obtained from wind-tunnel and flight tests.

This is yet to be resolved. Nevertheless, the drag-due-to-lift charac- teristics measured in flight agree reasonably well with predicted characteristics.

Extension of the lift-drag evaluation under full-scale conditions to determine scale effects and base-drag contributions will provide a better understanding of wind-tunnel results and allow interpretation of wind-tunnel tests of other advanced vehicles to full-scale conditions.

Heating In the research area of aerodynamic heating, a potential high-speed problem area, the highest temperatures recorded have been in the neigh- borhood of 400 ° F, obtained during a speed-buildup mission which resulted in a maximum Mach number of 3.2.

Figure 5 presents only a brief snmple of the type of temperature measurements being obtained. A sectional sketch is shown of the mid- span station on the wing with the chordwise distribution of skin tem- perature at the time peak temperature was realized. _his distribution occurred after the peak Mach nmmber of 3.2 for this flight was attained.

The lower skin temperatures are higher than those of the upper skin mainly because of angle-of-attack effects and partly because of the thinner skin on the lower _ur__ace. The lower temperatures near the leading edge are attributed to hhe heavy leading-edge heat sink designed to handle the large heating rates to be encountered _urlug the design mlssicas. An example of the internal variation in temperature is also shown in the plot in the upper right corner of this figure. The rapid rise in skin temperature produces a considerable lag and, consequently, differences in temperature between the free skin areas, spar caps, and internal webs. For example, at the _O-percent chord there is a differ- ence of about 900 ° F between the lower skin and the center of the web.

Present indications are that predictc_l and measured ful_-scale tempera- tures are in fair agreement, aud, therefore, a degree of assurance for e_xtenddng the f!i_ht pro_ram hag been p_T_v!d_,___ ee ee e--_*e eeeeeee,°**_ ,,st e .... s • :: • :: • :: 79 tl ! |)J )e le As mentioned, the data shown constitute only a brief sample of the temperatures measured, for there are 650 thermocouples located on the X-l_ to provide a rather complete coverage of skin and internal temperatures.

Future flights are planned w_th more nearly stabilized flight conditions to provide nondimensional heat-transfer coefficients which will be useful for applying results to other airplanes. Although plans have been made to obtain data on the X-l_ in the speed range up to M > 6, it might be added that this full-scale heating information is sorely needed for development of even the M = 3 to M = _ airplane.

Dynamic-Loads Problems H One of the problem areas which became evident early in the X-15 flight program is panel flutter. A review of the X-15 _tructural design shows that the type of structure and the materials utilized in the construction of the X-15 were governed by the heat environment anticipated during various hypersonic flight missions. The X-19 side- fairing I_nels were construc_ed of a flat sheet stiffened by a corru- gated backing, and no adequate analytical methods were, or are, avail- able to predict flutter of a flat-sheet paneS, much less the complex panels used on the X-I_. Moreover, prior to flight testing, no wind- tunnel tests had been performed to investigate panel flutter on these specific panels. During flight tests, panel flutter of the side- fairing panels was experienced. Subsequent wind-tunnel and flight tests provided a simple fix that appears adequate to avoid this phe- nomenon, at least for the present.

Figure 6 shows an example of the relative panel response measured during flight for the original panels and for the stiffened panels as a function of dynamic pressure. The upper curve represents response of the unstiffened panel, and the lower curve shows response of the stiff- ened panel. The abrupt increase in panel response for the unstiffened panel represents the start of panel flutter. The beneficial effect of the modification is illustrated by the general reduction of panel response and the absence of panel flutter. Flutter of the vertical-tail panels has also been detected during wind-tunnel tests, and modifications to the vertical-tail structure have been incorporated. Further _-ind- tunnel tests are planned to clear the airplane flight envelope to the design dynamic pressure, and continued monitoring b_ means of flight measurements and inspection during the flight program, particularly in a high-temperature and high-dynamic-pressure environment, will provide additional background for fu-.:ure advanced designs.

Inasmuch as recent general studies and experiences have indicated that panel flutter is the type most likely to be encountered in advanced designs and in a higher speed environment, it appears that additional studies are required to establish design procedures for avoiding this phenc._enon. _ _. , 8O • 4 _'" Recovery and Landing _other problem encountered In the X-15 flight research program pertains to the recovery and landing. The increasingly critical nature of the approach and landing maneuver as lift-drag ratios have decreased has caused a general focusing of attention and research effort on these problems as related to more advanced hlgh-speed aLrcraft and also to hypervelocity and reentry vehicles. For these advanced vehicles, it was thought that a lower limit of lift-drag ratio existed beyond which it was not possible to effect a safe landing. Landing of the X-l_ was even considered questionable.

During the past few years, the Flight Research Center has devoted much effort to evaluation of suitable approach and lauding techniques for vehicles having peak lift-drag ratios approaching values as low as 3.

As a result of these studies, which included flight simulation with air- craft such as the F-102 and F-104A, the approach and landing technique for the X-15 was evolved. A summary of the flight touchdown conditions experienced thus far is shown in figure 7 which presents the variation of vertical velocity with angle of attack. The touchdown vertical velocities and angles of attack have generally been well within the design envelope showa_ and the technique utilized is deemed satisfac- tory. As expected, with this technique a high level of pilot profi- ciency is required for landing, and increased pilot experience gener- ally provides improved approach and touchdown conditions. Inasmuch as future advanced military vehicles probably will utilize power and hence will have peak llft-drag ratios for landing which are grea_er than that of the X-15, no significant landing problems are foreseen. Howe_Tr, in the emergency power-off condition, where lift-drag ratio may be quite low, these advanced vehicles may benefit from the lauding techniques developed with the X-15. In addition, the X-15 has provlded, and will continue to provide, some significant advances to the state of the art for skid landing-gear systems.

Systems Evaluation In addition to the research performed in the areas discussed, information and experience are being obtained in preflight and flight evaluation of the major systems listed as follows: (I) stability aug- mentation system, (2) reaction controls, (3) adaptive controls, (_) controllers, (5) display, (6) inertial platform, (7) physiologlcal_ (8) operational, (9) hot nose, (lO) rocket engine, and (ll) energy management. Although most of these systems are being independently developed in various _ound-based envirom_nts, it is only in a flight environment, in combination vith other systemsp that a re_listic demon- stration can be effected. In contrast to the agreement found between flight 8nd predicted aerie characteristics thus far, the systems • _m Ii • oo o ee • • _ Pob _I experience in flight has not always been satisfactory. This _s more or less anticipated when a system is in the development or checkout stage.

In some instances, component reliability or integration problems have been encountered, and in others component or system development is required. Also_ some of these systems have never been flight evaluated and are being developed in laboratory or mockup studies for future flight use. In all cases, continuous product-improvement effort is, and will be,necessary to provide satisfactory flight operation and reliability. It is relatively certain that many of these systems o_" modifications of these systems will have future application to the vari- ous flight areas covered by the X-15, as well as to other flight regions.

Therefore, it is safe to say that these systems are, and will be, devel- oped to the satisfactory stage and should he available for application to other advanced vehicles.

CONCLUDING RF_ A discussion has been presented of the flight-research objectives and some of the results obtained to date in the X-15 flight-research program which are pertinent to the development of advanced military aircraft. Flight studies performed thus far, up to a Mach number of approximately 3, indicate that the aerodynamic force and heating data oltained agree reasonably well with wind-tunnel predicted data. More significant data, having an influence on other advanced vehicles, will be obtained in a number of research areas when the speed range is extended to Mach numbers between 3 and 6.

_ome aerodynamic problems, dynamic structural problems, and systems operational problems were encountered, and probably will continue to be encountered, as a result of the use of new configurations, materials, structure, and systems concerning which little or no practical knowl- edge is available or which require full-scale verification. These are some of the same problems which will be encountered in the development of any advanced aircraft s_d can be evaluated and finally solved auly by a realistic study in a flight environment. Therefore, the current flight studies being performed with the X-15 will provide information which will benefit the development of advanced military aircraft.

IIIIi'-----'- .° I • ........ . • • ::'" °.._ 8?

..... : '::."

• e,, I | ! ii

X-15 PERFORMANCE ENVELOPE

600xlO 3 \ \ \ 4OO \ \ ALTITUDE, FT % O, 20C XLR99 r ENGINE Iz I *1 I I I | I I 0 2 4 6 8 M Figure i X-15 STABILITY DERIVATIVES = =f4%10" ,016, .16 o FLIGHT .ol 2 = WIND TUNNEL _I) 0 .12 C-'Na, P_t DEG PERDEG 1)4 i i | I I I I I I i C

-=[ _o_[ o

_,._=[ _ \ c1_ oL _, -

_X__ 0 | 2 3 4 5 6 7 o i 2 5 4 5 t M M Figure 2 / !

_L

Jl s leo t eel ee ! o o o • 1; :,(-15 CONTROL-EFFECTIVENESS DERIVATIVES a = 0"-'15" ,._I_t m_ WIND TUNNEL CmBh, PER DEG -01 --'03 t-.02 _....._ ". FLIGHT Cnsv ' PER DEG =002 I 0 / , t I i i i I i %°. PERDEG .00_ "o_"_-'_.L.

i I I J I I i 0 I 2 3 4 5 6 7 M Figure 5

)(-15 DRAG CHARACTERISTICS

POWER OFF FLIGHT ...... WIND TUNNEL

// /

.4 _.9,.,M.,., /..2.c,/; ., CL .2

i--

I, , ' | 1 .. l 0 .08 .16 0 .08 .16 0 08 .16 CD CD CD Figure

X-15 WING TEMPERATURES

MIDSPAN STATION

_

_-o-- LOWE_ SK,N _.t 3/

6O0 , , _ V TEMR, eF .___._-- .l]|']j|l'lZlWll-lr_-_ :(: ; J J , * • . - TEM.P,_: 2O0 I I,, I I I I I I • I I ,I 0 20 40 60 80 100 % CHORD Figare

X-15 PANEL RESPONSE

1.0 RELATIVE PANEL RESPONSE .5 I. I I I I I J I, I Z00 400 600

8o0 tpoo

q, LBS/FT _

Figure 6

2D I 8_

X-15 TOUCHDOWN CONDITIONS

,,j 3 12 ,-I !

I0 GEAR DESIGN LIMIT _0 % 0 0 \ VERTICAL VELOCITY, FPS 6 O0 OGDO

o\

0 0 0 o % II_l I 4 6 8 I0 12 14 C 2 a, DEG Fig_ _e 7 I',4" VX. HI'PERSONIC CRUISE VEHICLES By/John M. Swihart and John R. Henry : j t , Langley Research Center .

N67-33076

IFfROI_JCTION Interest has been recently expressed in the military use of hyper- sonic cruise vehicles as b_nbers, transports, reconnaissance aircraft, and perhaps recoverable boosters. The purpose of this paper is to give a very brief summary of the state of the art in aerodynamics, aero- dynemic heating, propulsion, and to indlc_te the performance possibil- ities of _--_ch an aircraft based on this state of the art. A _eat deal of research applicable over broad regions of the hypersonic speed range is in progress (refs. i and 2, for example) but for the present purpose, discussion will be confined to cruising vehicles which are boosted to cruise speed and altitude. For this purpose a Mach number of 6 has been chosen as the design point, which is considered to be about the limiting Mach number for the use of hydrocarbon fuel and is also considered to be a logical starting point for the use of liquid hydrogen fUel.

S_BOLS inlet capture area afterbody drag coefficient CD, a drag coefficient at zero lift CD, O lift coefficient

%

pitching-moment coefficient D inlet capture _er

Do

F tnternat thrust L lift M n_ber _ __ ,:L-r ....

...... .. . _.

• e o • • * • _ • • e,e • • • -q -q • • -_q _q _-,,D 4e q dynamic pressure R Reynolds number transition Reynolds number gross weight angle of attack boattall angle C emissivity inlet kinetic energy efficiency nozzle kinetic energy efficiency _KN wlng-leadlng-edge sweep ALE RESULTS AND DISCUSSION Figure I illustrates a configuration concept for a hydrocarbon- fuel vehicle. This relatively high-fineness-ratio high-fuel-denslty model has a 70 ° single-wedge-slab delta wing with a thickness ratio of 2 percent. The fuselage is shown mounted on top of the flat-bottom wing; however, by interchanging fuselages and wing camber positions, several models were obtained and were investigated. There were no engine nacelles mounted on these models.

Figure 2 illustrates a configuration concept for a liquld-hydrogen- fuel vehicle. It has the same kind of wing as the hydrocarbon-fuel model, except the lea_Ling-edge sweep has been increased to 74 °, the wing tip _ms been clipped, and the very large fuselage necessary for the iow- density kvdrcgen fuel has been placod sy_metrieall_ on the wing.

A calculated drag breakdown for these two designs at a Mach number of 6 is as follows: • • • • OOI | n e@ coo 4@ 6ao • s oR In Component drag coefficient, percent CD, 0 T_pical Design Tall Body Wing Skin CD,O L friction Wave Base Wave Base Wave Base Hydrocarbon 4 5 1 2 66 2 2O fuel 0.0050 Hydrogen 2 1 30 16 _3 0.0056 fuel !

Typical values of full-scale minimum drag coefficient CD_ 0 are shown for each configuration and the contribution of the body, tail, and wlng to the minimum drag are listed as percentages. This table serves to illustrate that for a hydrocarbon-fuel design, the skin-friction drag is all important amd no roughness drag above the level for turbulent boundary layers (which has been discussed previously in part III of this volume) caube tolerated. The skln-friction drag fcr the hydrogen-fuel design is still very important; however, the wave drag associated with the large fuselage has become an appreciable part of the minlmumdrag.

For botlrconflgurations, the base drag for these slab wings, tails, and bodies is quite high.

The effect of boattailing on afterbody drag is shown in figure 3- These data are for a Ma_hnumber of 6 with a fully turbulent boundary layer, and the afterbody drag coefficient obtained by integrating the pressures along the afterbody and across the base are plotted against the boattail angle. The data were measured on a two-dimensional slab wing and the sketch "_ figure 3 shows how the model was boattailed with the base dimension corresponding to a boattail angle of 12 °. For the flat base, _ = 0 °, the drag is nearly equal to the estlm_ted vacuum value of the pressure coefficient, or (-l/M2); however, as at lower Mach numbers, substantial gains are realized from boattailing and a 33 percent re_luction is shown in CD, a by incorporating a boattail angle of 6 °.

Figure 4 shows the aerodynamic characteristics for the high-density hydrocarbon-fu_l model. These data were obtained in the Langley Unitary Plan Wind Tunnel at a Mach number of 4.63 and a Reynolds number of 7 × 106 based on the mean aerodynamic chord of the wing. Figure 5 shows data obtained in the Langley _X)-MPH 7- by 10-foot tunnel at a Mach num- ber of 0._5 and a Reynolds number of 2.5 × 106. The data in figures and 5 are for a configuration with the fuselage below and above the delta wing untri_ and with no vertical tail. The L/D is slightly higher vlth the body above than __bod_elow the wing, but the

90 ........ : • "'. "°. .'_

• t .. ." : ..... . . • l _ I I I I I I I I I I l I I I I I I I I I I I I • I I i configuration with the body below the wing _ives a higher value of Cm at zero lift. Many methods have been devised in the past to reduce the trim drag that would be associated with the body above the wing; and one of these, nose cant; for example, could be applied here. Shifting the body position changes the lift-curve slope slightly as would be expected and changes the position of zero lift by about ±l°. An impor- tant point to be noted in figures h and _ is that the data appear to be very smooth and show no radical nonlinearities, either at subsonic or high supersonic speeds.

L Figure 6 has been prepared to give an indication of the magnitude.

I of the llft-drag ratios that have been obtained. Solid and dashed lines are calculations for flat plates at a Reynolds number of 90 x 106. The .I transition Reynolds number was assumed to be zero (leading edge) for the solld-line curve and lO x 106 for the dash-llne curve. This value of transition is probably as high as can be expected, inasmuch as high values of leading-edge sweep tend to have a detrimental effect on laminar flow. (See part III.) Some preliminary data obtained at a Mach number of _ with a 70 ° delta wing indicate that transition occurred at a Reynolds number of about 5 x lO 6. These flat-plate data set an upper limit to be expected on L/D sa_d show a value of L/D of about lO at a Mach number of 6. The model data shown are not for complete models in the usual sense in that they have no engine nacelles; however, in a subsequent part of this discussion, the nacelle external drag has been subtracted from engine thrust, and the interference drag is believed to be very small. The data indicate that for high-denslty hydrocarbon-fuel vehicles, values of L/D of about 8 can be obtained at a Mach number of 6, whereas for low-denslty liquid-hydrogen-fuel models the values of L/D are near 6 or less over the speed range up to a Mach number of 10.5.

This level of aerodynamic performance is shown subsequently to be high enou@h to provide desirable ranges for cruising vehicles.

Figure 7 is concerned with the aerodynamic heating of the basic vehicle structure. The equilibrium surface temperatures calculated for a Mach number of 6 with an emissivity of 0.9 are tabulated for the wing of the hydrocarbon-fuel vehicle at an angle of attack of _o. The wing leading-edge diameter is I inch. The tabulated values indicate that the leadlng-edge temperature would be about 1,7_0 ° F, that the lower- surface temperatures from i_mediately in back of the leading to the trailing edges would range ___ 1,200 ° F to 1,_O0 ° F, and that the upper- surface temperatures would range from 700 ° F to 900 ° F. The leading edge might have to be made of some refractory material; however, the rest of the vehicle could probabl,v be constructed of one of the super alloys such as Ren_ 41, and perhaps the upper surface of the wing and also the bo_y, if mounted above the wing, could be constructed of titanium. Construction of such a vehicle is consequently believed to be within the capability of the aviation industry at the present ti_e.

. , s _ !

The propulsion systems for a vehicle of this type are next to be ccnsidered. The propulsion systems discussed are limited to the con- ventional ramjet type of engine. Other engine types are under study by various organizations throu6hout the United States; however, these other types are intended to operate either over a range of flight Mach numbers or at speeds substantially higher than a Mach number of 6.

Figure 8 presents a sketch of the propulsion unit under considcra- tion. It consists of three principal parts: the hypersonic inlet, the subsonic combustion chamber, and the exhaust nozzle. Much research has yet to be done on the hypersonic propulsion unit. Wit. regard to the inlet, in addition to the usual determinations of optimum configurations for missions of interest, questions and problems exist relative to the effects of boundary-layer cooling on the inlet performance, the method of handling the cooling, the design of structural and variable-geometry configurations, and the optimum compromises between boundary-layer bleed and total pressure recovery. Essentially no information is available on combustion chamber design, for example. Much information is needed on recc_blnation rates for the gases flowing through the exhaust nozzle and the associated effects on exhaust-nozzle design requirements. The ma_ter of the appropriate combination of special materials and cooling arrangement to be used for the diffuser, nozzle throat, and combustion chamber is under study. Recent research (ref. 3) by Connors and Obery of the _ASA Lewis Research Center indicated that the heat-transfer rate at the nozzle throat of a ramjet engine would need to be about 400 Btu per square foot per second in order to maintain a wall temperature of 1,_OO ° F. This coolin 6 rate is approximately one-fourth that being sustained in rocket engines for short duration; therefore, Connors and Obery concluded that it should be possible to cool ramjet engines for indefinite Periods of time. For the hypersonic ramjet engine to be really promising, a high-energy fuel with a large heat sink and no coking problems is required. Ordinary hydrocarbons do not fulfill these requirements and liquid hydrogen has many logistic disadvantages.

However, all of these problems which have been listed appear to be sub- Ject to solution with sufficient investment in research and development.

The proportions of the exhaust nozzle indicated In figure 8 were determined fr_n an analysis using real-gas Mollier diagrams (ref. 4) and three-dimensional characteristic computations for the external nacelle drag. The dimensions given apply to either hydrocarbon or hydrogen fUels because the proportions were nearly the same for the two fuels. The exit diameter of I.SD 0 corresponds to an underexpanded nozzle and represents the maximum thrust-minus-drag configurations. As the nozzle diameter is increased beyond this value, the external drag nozzle exit to throat diameter corresponds to an area ratio of 13.8, an exit Mmah number of 3.3_ and an exit static-pressure ratio of about 2.0. Hypersonic ramjet engines have been designed in the industry for ..... ..... ........

J e 6 • • 92 ........ _ " ,, ::. - s, t o • I e internal pressures up to TOO pounds per square inch absolute; therefore, the combustlon-chamber stagnation pressure of 98 pounds per square inch absolute for this unit should incur no additional structural pr_;ulerm.

The inlet stagnation temperature of 2,600 ° F and the combustion tempera- ture of 4,800 ° F will pose cooling problems and special materials, and perhaps regenerative cooling will be required. In this regard liquid hydrogen possesses ideal cooling properties and studies have shown that liquid hydrogen will provide sufficient cooling up to Mach numbers of 8 or perhaps 9- Other fuels such as frozen methane also offer possible solut ions.

In relation to the hypersonic-inlet problem, fi&-are 9 contains sketches of two inlet types which are under active consideration. The sketch at the left of the figure represents a three-dimensional all- external-compression splke-type inlet. The spike consists of a conical tip followed by an isentroplc compression surface, and boundary-layer bleed may or may not be used on the shoulder. Because of the local flow inclination the cowl lip may be set at a fairly high an_-le co avoid a strong reflected shock. This design will incur appreciable external drag and the thin annular throat may be subject to special structural and cooling problems. The great advantage of this inlet is that no variable geometry is required because no appreciable amount of internal compression occurs. The sketch in the right-hand half of the figure represents the same general type of inlet; however, the high cowling drag has been eliminated by reducing the external compression and sub- stituting internal compression. This effort to attain increased thrast minus drag results in a variable geometry requirement for starting the inlet. The United Aircraft Corporation has investigated both of these inlet types (refs. 5 and 6) and the NASA Lewis Research Center has tested the all-external-compression inlet (ref. 7)- Some representative data on these inlet types and on two-dimensio:n_i research inlets are presented in figure lO.

Total pressare recovery is given as a function of free-stream Mach number in figure lO, and curves of constant inlet kinetic energy effi- ciency based on real air computations are superimposed on the plot. The key in this figure indicates the source of the data: The United Aircraft Corporation data (labeled UAC) are from references 5 and 6. The Langley data are frc_ an unpublished work by John R. Henry, Lowell E. Hasel, and Ernest A. Mackley of the Langley Research Center, which was presented at a classified session of the SAE National Aeronautical Meeting (New York) in April 1960. The Lewis data are from references 7, 8, 9, and an unpublished investigation conducted by L. E. Stitt and D. L. Chubb at the Lewis Research Center.

The type of inlet ranging from all-external to all-lnternal com- pression is listed aud whether or not the flow field Is two dimensional (2-D) or three dimensional (3-D). The solid symbols are for fixed-geometry 9_ 00 e • B 0 oo o, • •o Io• • I 4_ .. • • I o P- inlets and the flagged symbols represent inlets with no boundary-layer bleed. For the cases with bleed, the measured recoveries were reduced by an amount appropriate to the excess drag associated with the bleed flow in order to obtain a true comparison between data with and without boundary-layer bleed. This adjustment to the data was accomplished on an equal thrust-minus-drag basis. The principal conclusion to be drawn from this figure is that inasmuch as the bulk of the data correspond to kinetic-energy efficiencies of 92 percent or higher, an assumption of 92 percent for computations of net thrust, range, and performance is very reasonable.

The engine-nacelle net-thrust coefficient based on capture area is presented in figure ll as a function of inlet-kinetic-energy efficiency for stoichiometrlc m'xtures of hydrogen and hydrocarbon fuels, w_ne drag component of the net-thrust coefficient includes both the pressure drag and friction drag for fully turbulent flow on the external surface of the axisymmetric nacelle. In addition, the drag coefficient was increased by 0.05 as an allowance for a drag increment due to the rounding of the cowl leading edge in order to maintain a temperature of 2,000 ° F or less.

Mollier diagrams for real air and gases were used in making the computa- tions (ref. 4). No attempt was made to determine optimum cruise equiv- alence ratios; however, other studies have shown that the optimum values are probably somewhat less than the value of 1 used in this analysis.

A nozzle-klnetic-energy efficiency of 0.975 was assumed in the computa- tions. The ticks and numbers appearing on the two curves give the values of specific fuel consumption associated with the particular thrust coefficients.

The primary purpose of thir figure is to show the general level of thrust coefficient and specific fuel consumption obtainable for each of the fuels at the inlet-kinetic-energy eff!cicncy of interest, 92 percent.

The thrust coefficient for _drogen is 0.98 which is only ii percent higher than the value for hydrocarbon fuel of 0.88; however, the specific fuel consumption of 1.19 for hydrogen is only 58 percent of the value of 3.09 for the hydrocarbon fuel. This advantage of hydrogen, of course, is offset to some extent by its low-density high-storage volume requirements.

The values for the aerodynamic characteristics of hypersonic con- figurations which have been presented earlier hereln have been combined with the propulsion umit-performauee values Just presented to give the range-payload-mission potentialities. The ranges have been computed by using the Breguet range eq_tion and do not include increments of range obtained during the boosted portion of the flight or the glide letdcwn with previous work on hypersonic glide vehicles amd _th ._ndustry studies on hypersonic cruise vehicles• 91; ....... _ • Q • • • _ • • i o • - • • • • ee eo • e Ill se i i a Ig The results of these performance computations are given in figure 12 which presents range in nautical miles as a function of payload for assumed gross weights of 200,000 and 290,000 pounds for both hydrogen and hydrocarbon fuels. The curves shown should be regarded as approxi- mate indlcation6 of the level of performance obtainable, inasmuch as no detailed design work ims been done on these configurations. It shc Jld also be noted that the weight of the solid propellant booster required to lift these vehicles to a design cruise speed of Mach number 6 at an altitude of approximately 80:000 to 100,_ feet _s roughly twice the weight of the vehicle itself, _o that for the 200,O00-pound machlne_ the L gross take-off weight of the cruise vehicle and its booster would be i about 600,000 pounds. This booster weight could be cut in half by boosting o__y to a Mash number of 3.0 and paying the weight penalty of i the varlable-geometry inlet required to makc the rs_et self-acceleratlng.

The booster weight could also be reduced by about one-half by incorporating an all-llquld system and using liquid hyd_-ogen as fuel• The _ydrogen fuel provides approximately 50 percent more range than the h_drocarbon fuel over the entire range of these calculations. This result shows that the low specific fuel consumption of hydrogen has out- weighed the adverse effect of hi@h-volume store@e requirements on the aerobic characteristics of the vehicle; however, it cannot be con- clu_ed that h_gen is best because of the ma_ logistics problems involved for military applications. Military missions of roughly lO,O00 nautical miles and 29,000 pounds of payload are of considerable interest sad the curves clearly show that this type of mission is obtaino_le with a boosted hypersonic cruise vehicle. Smaller ranges and payloads such as those applicable to reconnaissance missions could be accomplished with _,ch smaller gross weights than have been indicated here. It is possible that these smaller vehicle_ mould be launched from a recoverable booster such u has been under stud_ for Dyna-Soar and other space missions.

CONC_IOHB It is recognized that there are me_y pro1_lem areas for a b_personic cruise vehicle which will require intensive research for satisfactory solutions; however_ it appears that: l. The state of the art is such that llft-drmg ratios of sufficient mm@nltude to give satisfactory range can be obtained.

2. The the_cs of the propulsion units yield values of pro- pulsive efficiency which, when coupled with the serod_amie efficiency, _ _' '" !i. _ .i• " indicate desirable range-payload possibilities. Unfortunately research data on the materials and cooling methods for this engine are not well documented.

3-The aerodynamic heating of the vehicle is low enough to allow construction of such a vehicle within the present capability of the aviation industry.

_CES I. Armstrong, William 0., and Ladson, Charles L. (With Appendix A by Donald L. Baradell and Thomas A. Blackstock): Effects of Varia- tion in Body Orientation and Wing and Body Geometry on Lift-Drag Characteristics of a Series of Wing-Body Combinations at Mach Numbers From 5 to 18. NASA _M X-75, 1999.

I 2. Rainey, Robert W., Fettermam, David E. Jr., and Smith, Robert: i Summary of the Static Stability and Control Results of a S_personic Glider Investigation. NASA _M X-277, 1960.

i 5. Connors, James F., and 0bevy, Leonard J. : Some Considerations of Rlrpersonlc Inlets. Paper presented at 4th AGARD Combustion and Pro_ulslon Colloquium on High Math Number Air-Breathing Engines (Milan, Italy), Apr. 1960.

,. Hall, Eldon W., and Weber, Richard J. : Tables and Charts for Thermodynamic Calculations Involving Air and Fuels Containing Boron, Carbon, Hydrogen, sad Oxygen. NACA RM E56B27, 19_6.

9. Kepler, C. Edward: Performance of a Mach 4.0 Variable-Geometry Axisy_netric Inlet Having External-Plu_-Internal C(m_ression. Rep.

R-!285-12 (Contract NOa(s) 99-155-c), United Aircraft Corp., Sept. 1959.

6. McLafferty, George H. : Hype2sonic Inlet Studies at UAC Research Laboratories. F, ep. M-2000-113, United Aircraft Corp., Dec. 1959.

7- Stitt, Leonard E., and Flaherty, Richard J. : Experimental Investi- gation of a Mach 3 Isentropic Spike at sad Below Design S_eed.

HASA _4 X-4, 1999.

8. Connote, James F., and _nderson, Leverett A., Jr. : Performance of an Axisy_netric External-Cumpression and a Two-Di_.ensional External-Internal-Compresslon Inlet at Mach h.99- NASA Memo 12-18-98E, 1959.

9. Connors, Jam_s F., and Allen, John L. : Survey of Supersonic Inlets for High Mach Number Applications. NACA RM E_oO, 1958.

.__ n -,_- .

j-,, • 4b • • _01J s0 et a II0 $$1} I_ ! s u'% O rc_ !

i i ..,.,J I,.¢J • ¢b, _ Q

O

,g

I---

m

Z

I,¢J -,'-4 t,zJ I.L i ::1:: I %," , i el iN a • • • • . mA _" '''-_".: j('j ..._ ,._ ,_,. _.. -:_

.__

ce _i" hO .J UJ

Q

>-

\

h-

\ o .

(I)

Z

IJ

a

.b

ILl I,I.

I o, ,t "_.

_,._ i",:".--

. ,..: -_..; .. .; ._._._,., , - wn m 0 0 0 0 0 0

I--

• • • • • •

W

Q la..

L) U.

I.LJ

lO0 _.1 1.1.1o O_

_z

J

i m ro IN) L o oJ co _ o LI_

O_

_ tw

V--:_

TW OZ I Z_ i o ,_" o.

r_ F: , I_z W_ ) lOl 0oj ,,,-I P,, , ; • "-,,w ,! ,!.,, . : • . .

• Q , • • •

102 "':":" "'"'_i_li_.:!:i

AERODYNAMIC LIFT-DRAG RATIO UNTRIMMED I0 o.--- _ _.,e_CALCULATED /_')MA X MODEL R RT t-, FLAT PLATE 90xl06 0 I FLAT PLA_,E . IOXlO° l-J fO 0 H YDROCt,_;.=JON ,i 0 " HYDROGEN ,, 0 I ,,. I t .......... i 0 4 6 8 I0 12 M Figure 6 EQUIUBRIUM SURFACE TEMPERATURES M=6, ¢l=5", • =0.90 Fi_z'e "f m_

DIAGRAM OF PROPULSION SYSTEM

M=6.0

T

%

1.3D 0

I_%

m MACH NUMBER

3.3

98 98 .75

PRESSURE, PSIA ?.600 4800 TEMPERATURE, o F

HYPERSONIC INLET DESIGNS

SPIKE-EXTERNAL-INTERNAL SPikE-EXTERNAL HIGH DRAG COWL -7 LOW DRAG COWl..- 7 SPIKE BLEED-x,, / SPI_ STARTING l I ,l • e oe .. ..

'i!!i i !_i_i i

ti

!

I----.

PO 1 o - i-" ,_, _ °'%, , = , | b ! i RAM JET NET THRUST - M = 6.0 STOICHIOME T RIC, _KN ='975 I- _.0 NET THRUST COEFFICIENT, .9 • %D F-D q A o /--'FYPICAL _KI "'_| ! t I t l 0 90 91 92 93 94 95 INLET EFFICIENCY, _KI,% Figure ii CRUISE RANGE - PAYLOAD AT M= 6.0 _KI" .92 : e/KN ".975 --_, 250 I000 I0 NAUT. Nil : HYDROCARBON'_ -"' _ _ 250 e _200 m !

= '. ! I !

0 5 i0 IS _'ZO_ 25 PAYLOAD/IO00,LB YtSuz'e 12 iil ! :.!y !: io7 VII. SOME STRUCTUPAL AND MA_ERIAI_ COESIDE_IORS FOR MARRED MILITARY AIRCRAFT Eldon E. Mathauser_ Richard A. Pride, and Avraham Berkovits Langley Research Center INTRODUCTIO_ The design of lightweight, efficient structures is one of the funda- mental requirements for achievement of high performance in military air- craft. In this psper some of the structural and materi_is considera- tions that are important to the strength, weight, and integrity of aircraft structures are revte_md. A comparison of s_e structural materials is made on the basis of weight-strength and tear resistance.

In the area of structural design, the relative weights of several types of construction that are of current interest suT reviewed and the influ- ence of different materials and the effects of elevate_ temperatures on structural weight are indicated. Lastly, other fI_tors that are of impo_ce in structural design of future aircraft _re discussed. These include strength under ncnunifora temperatures, creep, sonic fatigue, and panel flutter.

MA_IALS Weight-Strength Cc=parlson The relative efficiency Of structural materials is frequently deterained on a veight-sta_agth basis by use of plots of the type shown in figure A. Relative weight is plotted _i_st texture for 7075-T6 slumAmm alloy, 6AI-_Y titanium alloy, PH 15-7 Mb stainless steel, Ren_ _I nickel all_, emd a material of e_iderable structural interest, beryllitm. This cemperison is _ade on the basis of density and ultimate tensile strength and provides a beads for selection of m_n_-weisht tensile umbers. For the selection cf _embers under ecsprusi.,_ Icsd_, mat_,_al properties such as Youn_'s _odulus and yield strength vottld be utilized.

• o_e that the mtainlems E+_eel and the titanium alloy are either ecmpetltiwe or e,_i_ welghtvise to the al_ alley at _ tern.

per_ture _nA that relati_l_ little vei_ht in_-ease is obt_LneA w_th then two _atorta_ for te_eraturu u_ to (_0 ° • or 700 ° F. At • • , - • ...

I08 ........

higher temperatures, a change to nickel alloy materials, such as Ren_ _i, appears desirable in order to conserve weight.

Weight-_trength considerations such as these.are of interest in the selectlon of st_actural materials_ however, other factors are becoming increasJ ngly important.

Tear Resistance L Tear resistance is of concern for transport-type alrcra_tbecause i of the possibility of catastrophic failures of aircraft in service. It is expected to be of concern when high-strength, thin-gage sheet mate- rials are used in the structure. I The tear resistance of some structural materials is presented in figure 2. (See ref. l.) The average- or gross-area failure stress _f for sheet containing a crack, divided by the ultimate tensile strength of the material _ULT' is plotted against the ratio x/b, where x is the crack length and b is the plate width. The line labeled "cQm- pletely ductile" represents a material that is completely insensitive to the presence of a crack. Among the materials shown here, 6061-T4 alu- minum alloy indicates the best tear resistance am=l _20 stainless steel the poorest Nute that several high-strength steels are superior on this bp_is to the 20e_-T3 and7079-T6 aluminum alloys in eurrent use.

Data on tear resistance ere not available for mam_ of the struc- tural materials of interest; and in some cases, the data cover onl,v a small range as shown, for example, for the PH 19- 7 Mo stalnless steel and 6KI-4V titanium alloy. Furthermore, very little elevate_-t_mperature data of th_ type are available. General agreement does not exist as to the slgul_ cance of this type of information, although it is recognized tha_ structural sheet should resist tearing either from cracks that develop slowly from fatigue or suddenl_ from penetration by foreign objects• Continuing efforts should be male Lo identify tear-resistaut structural materials at low and elevated temperatures, to staudar_ize test methods, and to establish the slgnlf_2auce of this type of data in structural design.

STRUCTURAL DESIGN _rpe of Con.t_action Consideration is next given to structural deslgn. Three types of construction thatm_Tbe of particular interest for alreraftwlngs are

shownin figure 3. The view on the left indicates honeycomb-sandwich con-

struction in which the sandwich panels are fabricated either by adhesive bonding for use at relatively low temperatures or by brazing for appli- cations at higher temperatures. Construction utilizing an open-face sand- wich panel is indicated in the view in the center. This sm_dwich panel consists of a single corrugated sheet welded to a face sheet and includes transverse stiffeners attached to the corrugated sheet. The view on the right shows a stiffened panel type of construction utilizing hat-section stiffeuers that are either welded or riveted to the face sheet. In all three types of constr_iction the longitudinal webs are corrugated ar_ in the stiffened-ranel design the transverse ribs are also corrugated. Cor- rugated webs and ribs would alleviate thermal stresses that are produced by differences between the temperatures of the upper and lower wl.ng sur- faces. All of these designs are characterized by thin-gage sheet and genersll_ close spacing of the supports. These designs also reflect fabrication complexity that is coupled with high cost, particularly for the brazed honeycomb-sandwich type of construction. The relative weights of these types of construction will be examined.

In fSgure 4 the relative weights of the three types of idealized wing constructions investigated are plotted against the structural or l,_sding index. The weight, divided by the square root of the bending m(_nt, is plotted against the bending m*Jment divided by the square of the wing depth. The honeycomb-sandwich wing indicates the leas% weight, the minimum weight of the stiffened-pare el wing is 15 percent greater, and that of the open-face sandwich wing 35 percent greater.

It is significant to note that for a given bending moment the minimum- weight stiffened-paneL and open-face-sandwich designs are associated with thinner wings than the minimum-weight honeyccmb-sandwich design.

In this plot wing designs representative of sQme of the current high- performance fighters appear at values of the loading index greater than 1. O, where weight is not sensitive to the type of construction but is rather directly dependent upon the yiel_ _trength of the material. The wing design of a proposed __upersonic bomber falls in the loa_ing-index range between 0.I and 1.0 where the curves indicate a minimum weight.

The relative weights indicated in figure 4 do not include the weights of attachments between webs and cover panels and do not include the weights of the brazing material stud reinforcements that are required with honeyccmb-samdwich construction. Sume of the indicated weight advantage of the honeycomb-sandwich construction would be nullified by the addition of these weights.

Structural Materials Next, the influence of structural materials upon structural weight is examiue_. _nis comp_'son is made in figure 5 for several structural • • L • •

Ii0

: • . ...

' • " • Q•o ee materials utilizing honeycomb-sandwich wing construction. The weight, divided by the square root of the bending moment, is again plotted against the bending moment divided by the square of the wing depth• Note that the alumlnum-alloy and titanlum-alloy designs are competitive weightwise, whereas the stainless-steel design is 40 percent heavier, and the nickel-alloy design is 7_ percent heavier• The significant point in this comparison is that the most efficient aluminum-alloy wing has greater depth than the titanium-alloy or stalnless-steel wing. This res"It is of particular interest because titani_Jm-alloy and stainless- steel designs are generally associated with high-speed aircraft that L require thin wings for high performance.

i I Elevated Tempezatures Consideration is given next to the effect of elevated temperatures on structural weight. The minimum weights corresponding to the lowest point on curves such ez those shown in figure 5 have been obtained over a range of temperatures for the indicated materials. These minimum weights are shown in figure 6. The relative weight of honeycomb- sandwich wings is plotted against temperature. The results are based on materials data for l, OOO hours of exposure at temperature. These weight results suggest that aluminum-alloy construction would be satis- factory up to approximately 200o F, titanium alloy and stainless steel up to 700 ° F or 800 ° F, and Ren@ 41 up to 1,200 ° F or 1,300 ° F. Above these respective temperatures_ which ms_ be taken as llmit_ng temperatures for long-time application, very rapid weight increases are obtained. Note that for each material shown only a modest weight change occurs between room temperature and the limiting temperature noted previously for long- time application• These weight changes are on the order of 10 to 20 per- cent. Greater differences exist between the wing weights at room tempera- ture for some of the indicated materials• T_e;3e results on structural design have indicated relative weights of idealized wings for several materials over a range of temperatures.

To date, this study has not been extended to cylindrical shells repre- sentative of fuselages over the complete range of materials an_ temper- atures indicated in figure 6; however, results approximately similar to those precented would be expected from such an analysis.

OTH_ FACTORS Strength Under Nonuniform Temperatures The weight c_nparisons presented in figurea 4 to 6 have been obtaine_ under the assumption that the temperature of the structure is Ill ee oee uniform. It is recogniz_d that aircraft structures will be :_ubjer',_d to nonuniform temperatures in PliEht, and for this rem;on a brier' di:;- .'+ n • , I cussion on the effects of nonuniform temperatures on __r._tur.L_ :;t.r._._l.

is pre_ented next. Experimental studies (refs. 2 and }) generally indicate that maximum strength is independent of thermal stresses that are induced by nonuniform temperatures, whereas the load for buckling and permanent deformation is reduced by the presence of thermal streu_;es.

Some pertinent results on sandwich plates are now exmnined (fig. 'I').

The average stress at maximum load of is plotted against the average temperature T for 17-7 Ph stainless-steel corrugated-core sandwiches.

_ne face sheets of these small sandwich specimens were heated to temper- atures T 1 and T2, s_ the specimens were then subjected to axial compressive loading to determine the crippling strength. Tests were made with temperature differences between the faces of 200o F, _00 ° F, and 600 ° F. _ne dashed-line curve indicates experimental strength of the sandwiches under uniform temperatures, and the solid-line curves are calculated strengths for the indicated face-temperature differences.

_e calculated curves were obtained from a summation of the strengths of the individual plate elements of ti_e sandwich at the respective t_mperature of es_h element. Experimental data were obtained over the indicated temperature range and were in agreement with these calculated curves. 1_nese results indicate that thermal stresses did not influence the maximum strer_Eth. Although maximum strength was not influenced significantly by thermal stresses in these tests, the importance of thermal stresses in initiating undesirable deformation should not be overlooked. Each structural design will require detailed analysis to evaluate the effects of nonuniform temperatures and further studies _ith emphasis on desig_n features that minimize thermal stresses but preserve structural strength and stiffness are of interest. An example of such a study is described in reference _.

0reep _"ne problem of creep at elevated temperatures has attracted con- sider_le attention during the past few years because of its assumed importance on structural design. Studies m_le by the EASA in the past have indicated little likelihood that creep will be a ma_or problem.

This conclusion was based on both analysis and upon experimental data obtained under constant load a_ constant temperatures.

This conclusion m_ be dra_n from figure 8. The required weight of a tensile member is plotted against temperature for three structural materials. _e solld-line curves indicate the weight required for strength based on ulti_te load after l, 6K)O hours of exposure to temper- ature. Ultimate load is assu_ to be 3-75 times the 1 g loscl. The

lip

shaded regions define the temperature range for each material where creep

maybecomea factor in structural design. The left boundary of each

shadedarea indicates the required weight for 0.02 percent creep strain

in i_ CO0 hours st 1 g load andthe right boundary, the required weight

for creep rupture under the same load and time. Note that the tempera- ture range where creep m_v become significant for each material is rather narrow. Furthermore, creep does not become a design consideration until temperatures are reached where the strength of the material dete- riorates rapidly. It thus appears that when temperatures are encountered in which creep m_ become a problem, it will generally be necessary to convert to a material suitable for use at higher temper ture for strength reasons, and the creep problem will be eliminated.

Recent stlxlies on stm_etural __ssemb!!es ut_lizlng vary!_n_ loads representative of load-time relations for present fighters and bombers have again supported the conclusion that creep does not appear to be a major stm_ctural problem for aircraft. One note of caution is offered.

These conclusions are based on relatively short-tlme results compared with the desired life of some aircraft, particularly trar_ports. Some additional work to extend the experimental work into longer times ms_ be of interest; however, it is believed such results will support the present conclusions.

Sonic Fatigue In the area of acoustic fatigue, the underlying cause of structural difficulties is tl _ use of increasingly powerful propulsion systems.

Considerable effort has been made to obtain a better understauding of this problem ar_1 to improve the . oise resistance capabilities of air- craft structures. Examples of some detailed _esign features that mini- maze noise-induced structural fatigue are presented herein. Methods for fastening the skin to the ribs to determine features that are resist- ant to sonic fatigue are first considered. In figure 9 are shown several skin-rib Joints (ref. 5). The top row of numbers represents the fatigue life in minutes for these various Joints at a 160-deeibel noise level.

The design shown at the right was also tested at a noise level of 170 dec- ibels. The design on the left consists of a sheet and rib stiffener that failed in 17 minutes at the 160-decibel noise level. Addition of a doubler strip shown in the a_Jacent figure increased the noise fatigue life by a factor of approximately lO. The addition of a second rib stiffener to improve symmetry further increased the fatigue life.

Finally, use of bonding rather than riveting to decrease stress concen- trations resulted in still further increased fatigue life.

Another form of construction that has been used successfully for many high-intensity no'_ze a_lications is the honeycomb-sandwich. In figure I0 a honeycomb-sandwich panel is shown at the top with possible D B i • _ • • • m L , D • , • • * t • • t - D IDS ,, 1L3

!ii : i

oo damage areas due to a_oustlc fatigue indicated. At the bottom are shown various possible schemes for attachment of these panels to the main stm_c- ture. Fatigue damage m_" occur at the attachment points, in the bend radius of the edge-former of the panel, and in the bond between the core and face sheet or in the cell walls themselves. Among the different edge treatments indicated, the crushed cell walls are not satisfactory from the fatigue standpoint because the crushed cells are prone to fatigue cracks. Specimens fabricated with a formed doubler at the edge as shown in the upper right-har_1 view have survived for 90 hours at a nuise level of 160 decibels. Other promising edge treatments for which results are L as yet unavailable include the metal insert and the densified core.

2 This brief study of sonic fatigue has indicated some problem areas I in structural design that will undoubtedly be of importance to future aircraft. In all probability the acoustic fatigue problem will become more severe as a result of use of high-strength, thin-gage materials, coupled with construction that will utilize large numbers of tlr_ weld Joints that are potential sources for fatigue cracks. More detailed discussion of noise problems associated with manned aircraft is contained in chapter VIII of this volume.

Panel Flutter Panel flutter has an important bearing on structural integrity. It is of particular importance for structural surfaces that are fabricated from thin sheets of high-strength, hlgh-density materials that are designed to carry small structural loads.

The panel flutter problem will be examined in terms of some experi- mental information given in figure ll. (See refs. 4 and 6. ) The panel- flutter parameter in the ordinate is a modified thickness-length ratio where tEF F is an effective panel thickness, L is the panel length, M is the Mach number, E is Youmg's modulus for the panel m:,t_.rlal, and q is the dynamic pressure at flutter. The parameter on %he abscissa is a length-width ratio where L is the panel length and BEF F is the effective width of the panel. An envelope curve has been drswn to enclose the upper limits of more than iO0 flutter tests on both flat aug corrugation-stiffened panels. The flutter region lles below %.his envelope curve.

The purpose of figure ii is to demonstra :-u the influence of corru- gation orientation relative to the airflow on panel flutter. Two tests a_-e singled out for consideration. Identical square panels fabricated from thin-gsge sheet were teste_ in the Langley Unitary Plan wind tunnel.

The panel shown at the lower right was mounted so that the airflow was perpendicular to the corru_,ation sxls. Panel flutter developed during

" • • • •

I14

"P • q ' _ t the test a_l the model was destroyed. The second panel was mounted so that airflow was parallel to the axis of the corrugations and at the same dynamic pressure showed no indication of flutter. On the basis of this flutter parameter, a PO-fold increase in dynamic pressure would be required to move this test point into the flutter region• The X-I5 research airplane has corrugation-backed fairing panels along the sides of the fUselage with the corrugations perpendicular to the airflow. Wind-tunnel flutter tests of this fairin_ panel, shown by the test point indicated that flutter could occur within the operating re-nge of the X-15.

Figure 12 shows the fairing panels on the X-19 as well as an enlarged view of the interior s_de of the panels. The panels consist of a flat outer sheet welded to an inner sheet that contains the corrugations.

Indications of flutter were obtained in _light tests. This flutter has been stopped or at least considerably alleviated by the sddltlon of a longitudinal stiffener riveted to the crests of the corrugations. It is of interest to note that fatigue cracks are developing in these particular panels. These fatigue cracks or__glnate at holes that were drilled o_ the crest of each corrugation near the panel ends to relieve gas pressure during heat treatment. These fa_ _ue cracks cohtinue to develop sub- sequent to the addition of the transverse stiffener. In view of these flutter and fatigue difficulties, it is apparent that continued efforts are needed to obtain further insight into these proElems s_d to define structural design.", that are resistant to flutter and fatigue.

CON_LUDI_ EEMARKS Several structural and materials proble_as that are of interest for manned militarj aircraft have been reviewed end pertinent anal_ical s_ experimental results have been presented. Further efforts in these prob- l,_ areas have been indicated in order to guarantee structural integrity and high performance in manned military aircraft of the future.

.LA • . ..: : • REFERENCES i. Melcon, M. A. : A Survey of the Structural Properties of Some IHgh Strength Sheet Steels. Rep. lOl, AGARD North Atlantic Treaty Organization (Paris), Apr. 1997.

2. Zender, George W., and Pride, Richard A. : The Combinations of _,erma] an_ Load Stresses for the Onset of Permanent Buckling in Plates.

NACA TN 4095, 1957.

5. Pride, Richard A., and Hall, John B., Jr.: Transient Heating Effects on the Bending Strength of Integral Alumlnum-Alloy Box Besms. NACA TN 4205, 1958.

4. Pride, Richard A., Royster, Dick M., a_d r _.s, Bobble F.: Experl- NASA mental Study of a Hot Structure for a Reentry Vehicle.

TM X-51h, 1960.

_. Edson, John R.: Review of Testing and Info_nation on Sonic Fatigue.

Structural Dev. Note No. 58 (Doc. No. D-17150), Boeing Airplane Co., Mar. 7, 19_7.

6. Kordes, Eldon E., Tuovila, Welmer J., and Guy, Lawrence D.: Flutter Research on Skin 2anels. NASA TND-451, 1960.

_;_.. ii i' i !_i"'" " .." . - ." i -.."

u.6 !

i-J IX) D • • e • ee e vl I o • • • • • • • e • . e eo 4 • • eeo ele ]- _'1' m

×!-o

N

o. _ _. _.. N. o

li8 I .!

Z

Z

r

U

11_) I I _ I • e I

Z

/

I

I

°. • !

F-'

'7'

f _ • D eoe • • • lie • ].2]_ _D m

Z _,-

]o

r-i ,-4 i o4 OU- _0

LG

o.," ::D _0 rr i,i O_ -,-4 l.iJ

01-

lid

! I

TO 0

i"

ill • "t.o,._2 + i+ _m _4 00s oQ nl s 0 I I | I : n s

c:L

r_

z

I--

::}

C_ .,.,.J

Z

a}

:::)

.,-,I F-_

<:>

0:) I--.

u I,,LI !

Z

I..-

{/)

41, qP r,,,,.

m

<:>

I',., m ! I 125 _'_iqU_UnilllilmD" J el) (P e0e • ooo 00 !

• : .. .: , . .; :- I I o • II • le6 el eto * lie | Qe - " " eeb ee • ! i ! L _ _i -_ i i 'i , e t | ! i I a ! i "! I - | I !

I I II ,-4 q-4 I l I ; , i • • • • • e ! D e • • • J D_ s ) s i i 9 • • • • OB_ _ _ . ! , i ! !

oR DPD O0 6' e • • IIO i ; I

w

F-

_J

U.

_J

W

Z

Z

m

u_

i

!

X

_.m L pp,ECEDING PAGE BLANK, NOT FILMED.

By Harvey H. Hubbard and Domenic J. Maglieri J L_ey Research Center

N67-33078

INTRODUCTION Past experience has indicated that the noise problems of an air- craft are closely related co its design and the manner in which it is operated. Thus, in order to assess properly the potential noise prob- lems of a future aircraft, a knowledge of tile main features of its aerocl_c configuration is required as well as an appreciation for the various mission profiles assigned to it.

The material of figure i has been taken from various configuration studies, and the values listed are thought to be representative of three future aircraft types to which the discussi_s of the pre_ent paper will be limited. These are subsonic propeller- and Jet-powered V/STOL air- cra_t, larp hlgh-altltu_e supersonic-cruise aircraft, and special mis- sion aircraft capable of supersonic flight at low altitudes. The values given in figure i do not apply, to ar_ specific designs but are believed to Be realistic for these various aircraft types. In this paper there will be no attempt to d_cument the noise problems anticipated for each of these aircraft completely, but rather the discussion will be limited to those problem areas that are inherently associated with each because of its design and the missions to be performed.

Figure 2 indicates the main sources o ¢ noise for each aircraft type.

For the V/S_OL type aircraft, the main noise sources are the power plants.

Adverse coamnmlty reactio_ to noise during take-off and landing m_y be a problem for all three aircraft and will be discussed specifically for the V/STOL mad auper_c transport. JLlthot_h sonic fatigue due to the power p_t$ _.l _t be covered in this paper, it should be pointed out _ere that the proble_ for the V/STOL aircraft are similar in nature to those for current aircraft. Plac_ poser plants in the rear of the airframe, as has been Imdlcated in mam_ In_x_ed s_ersontc-transport desi_s, will tend to minimise but not necesaaril_ eliminate the problem. For aircraft operstlng at hll_ 4_c _ress_es, bo_-la_er noise is the main _erm and v111 be disclosed for a rsm_ of opcrsting conditions of interest for both h_h-altltude and los-altitude supersonic aircraft.

_hoe_-vmve-motse _lems are, of course, om_r of concern for s_perso_Ic- fl14_t operstlons. These trLL1 be discussed from the standpoint of min- i: i_sin_ _ee and property da_e duriM routine supersen_c-flisht _tl -- -'-- _m 12 for special m111ta:7 mlUici.

ha, -- _oe • _30 ,_ 1: .... : .....

eo e41 • e-_ f_ e-e qJe POWER PLANTS Of particular concern in the operation of V/STOL aircraft is the possible adverse community reaction around airports due to power-pl_nt noise during take-off and landing operations. Estimates have been made of the noise characteristics of two proposed V/STOL aircraft in an attempt to evaluate their noise problems for commerclal-type operations.

Some of the operating rules assumed for these aircraft are illustrated in figure 5. It is assumed that they would operate from either eonven- t_ onal airports or short-haul terminals. All V/STOL aircraft are assumed to cllmbout at a lO° geometric angle. This cllmbout angle is maintained to an altitude of 1,900 feet at which point a transition is made to level flight. This level-flight condition is continued out beyond, the area of air traffic congestion before the climb to cruising altitude is made.

The reverse of this procedure is used during the landing phase to the point where the approach is initiated, and then a 6° geometric approach angle is assumed.

Estimated noise data for STOL take-offs and landings are presented in figures _ and 9. The data of the figures apply directly to the STOL conditions; the principal conclusions, however, also apply to V/STOL conditions except in th- areas close to the termlm,1. Perceived noise levels (FNdb) for the 7 _catlon along the ground track of the air- craft are plotted in figure 4 as a function of distance from the point of llft-off in miles. Also shown on the figure for comparison are available data for conventional four-englne transport aircraft (ref. 1).

It is assumed that the STOL aircraft has two turboprop engines or two turbofan engines. The horizontal llne of small dashes in the centex of each figure corresponds to an acceptable noise level in some communities for daylight and early evening operations. Note that levels below the llne are considered acceptable whereas levels above the llne are not considered ac_.eptable. The main objective is to operate in such a way that the perceived noise levels on the ground become equal to or less than the acceptable level in as short a ground distance as possible.

It will be noted from figure h that the propeller- and Jet-powered STOL aircraft achieve acceptable noise levels in a shorter distance from llft-off than conventlomal transports. These reductions result mainly from the different noise spectra and the steeper cl/mbout capa- bility of the 8TOL aircraft.

Similar data are presented for landing in figure 5. The obJe:_Ive in landing is to operate the aircraft so that the noise levels remain at acceptable _lues within as short a distance as possible from the point of touchdown. It can be seen that at a given distance from the point of touchdown the perceived noise levels associated with propeller- driven STOL aircraft are somewhat higher than those for the conventlonal propeller transports. This increase in noise level is mainly due to L_L -. .. -_ .......

"_ • le

!!i

"" i: : .'.. :,.''..° • the higher power settings required during landing. The jet-powered STfOL aircraft has lower noise levels than current jet transports because of different noise spectra and a steeper approach angle.

The airspeed of the STOL aircraft during landing is considerably lower than that for conventional aircraft, and hence the duration of noise exposure for an observer on the ground is proportionately longer.

There is, thus, a possibility that the acceptable noise level for V/STOL type operations would tend to be lower than the acceptable level for conventional airplane operations. Also of concern is the fact that STOL operations may be carried on at local terLinals in the vicinity of which the noise tolerance may be less than in more densely populated areas. It should be noted that the STOL aircraft in landing at a con- ventional airport would probably touchdo_m about one mile closer to the center of the airport than the conventional aircraft. Hence, its noise at a given distance from the end of the runway would generally be lower than that for conventional transports.

There is a simil_r concern for the community reaction to supersonic- transport-type operations near conventional airports. Cllmbout will be made at about a l0 ° geometric angle, and the approach path will be at about a 5 ° geometric angle.

In figure 6, perceived noise levels are plotted as a function of horizontal distance in miles from the point of lift-off and touchdown.

Comparisons are again made with available data for conventional transport-type aircraft which are indicated by the hatched areas in the figure. For the take-off condition, the turbofan-powered super- sonic transport (SST) is seen to have lower perceived noise levels than the current Jet-transport aircraft. This results mainly from the fact that the larger thrust-to-weight ratio of the supersonic transport makes it capable of a steeper cllmbout angle.

For the landing condition in which the approach angles are about equal to those presently being used, the estimated range of perceived noise levels for the supersonic transport are shown by the crosshatched area in figure 6. A range of values is included because of the uncer- tainty in evaluating the airframe noise component. If noise from the airframe were not significant, then it is believed that the overall perceived noise levels during landing could be reduced to values near the lower extremity of the crosshatching in figure 6.

If turbojet engines with noise suppressors were used, it is believed that the obtainable perceived noise levels would only approach those of the upper extremity of the crosshatchzd area in the figure.

this type of aircraft is a formidable one because of the requirements for variable area exits and retraction during cruise flight.

ee e_ • _ UQ_ _9o • NNII_RY-_YER NOI_ On current airplanes the boundary-layer noise is mainly of concern from the standpoint of passenger comfort. For future aircraft, parti- cularly those capable of supersonie-fli@ht speeds, there is a concern not only for passenger comfort but also for noise-induced damage to the skin structure of the airplane.

Results from a large number of experimental investigations have L suggested that the boundary-layer-noise pressures on an aircraft sur£ace i are roughly proportional to the local dynamic pressures (ref. 2). A brief summary of existing information relating to surface-pressure levels I is given in figure 7 for a range of dynamic pressures of interest for future aircraft. In this figure, the term V_ z is the mean square of fluctuating pressure and q is the dynamic pressure. Experiments at subsonic speeds have indicated that the noise pressures are approximately equal to 0.006 times the dynamic pressure. This relationship is illus- trated by the solid llne in figure 7. Recent wlnd-tunnel and flight tests have indicated that the empirical constant in the above relation may be as low as 0.002 at supersonic speeds. This difference is equi- valent to about lO db, as indicated by the hatched area below the solid llne. On the other hand, recent measurements from Project Mercury space vehicles have indicated that in regions of separated flow the surface- pressure levels may be as high as 10 db above those indicated by the solid line. (See figs. 8and 9 of ref. 5.) This increase due to flow separation is indicated by the crosshatched area above the solid line.

Thus, at any given value of dynamic pressure, a wide range of surface- pressure levels may exist depending on the flow conditions. Shown also on figure 7 is a horizontal dashed line at a surface-pressure level of 140 db. It is believed that pressure levels higher than this may, under some conditions, cause structural damage, i_he locations of the ticks and the sketches at the bottom of the figure indicate the approximate maxlmumdynamlc-pressure values associated with each aircraft type. It can be seen that the STOL type aircraft will probably encounter little, if any, damm_e to the structure because of boundary-layer noise. On the other hand, the supersonic transport and special mission aircraft will probably have large areas of surface structure overwhich the pressure levels are sufficiently high to cause damage.

Further considerations relating to the boundary-layer-noise prob- lem are illustrated in figure 8. Boundary-layer thickness and surface pressures are indicated as a function of distance along the airplane fuselage. At the front of the aircraft, there is a region of laminar flow in which the surface-pressure levels are relatively low. Where the shading begins there is then a transition to turbulent boundary layer a short distance back along the fuselage and this turbulent

L

) boundary layer thickens up to,_rd the rear the aircraft,. Measurements have suggested that the overall fluctuati_g surface-pressure magnitudes are essentially constant along the fuselage although the spectrum shape varies considerably with boundary-layer thickness. As a result of this spectrum change, the surface pressures vary as indicated schematically in the bottom part of the figure. Near the front of the airplane, where the boundary layer is thin, the high frequencies predominate and the low- (audible) frequency pressures are small. Toward the rear of the aircraft where the boundary layer is thicker, the low frequencies pre- dominate and the high- (ultrasonic) frequency pressures are small (ref. 2).

The problems of acoustic fatigue of the skin surfaces and noise insulation of the interior compartments both involve the dynamic responses of the structure which are usually in the audible-frequency range. Thus, it would appear that both the acoustic-fatigue problem and interior-fuselage noise would be more troublesome in the aft portions of the aircraft.

SHOCK-WAVE NOISE Additional sources of noise in the operation of supersonic air- craft are the shock waves which result in sonic booms. Although these resulting sonic-boom disturbances may be observed throughout all super- sonic phases of the flight, the mDst serious problem._ appear to be associated with the climb phase where sonic booms may be produced at reduced altitudes (ref. h).

The material of figure 9 suggests an approach to solving the sonic- boom problem for the supersonic transport during the climb phase. The hatched area represents combinations of Math number and altitude which may result in damage to structures on the ground. The shaded area above the hatching represents combinations of Ma,-h n,/m_r and altitude for which sonic booms -will be observe_ on the gro.md and which may be annoying but will not cause damage.

The main objective in this flight operation is to travel from ground level to cruise conditions without intersecting the damage area.

This may be accomplished by climbing subsonlc_.lly to some intermediate altitude, accelerating to supersonic speeds i: level flight, and then finally climbing and accelerating to cruise c:.nditions. This altitude of 35,000 feet is considered an absolute minimum value and for a large airplane should probably be in the vicinity of hS,OO0 feet.

There has been some concern about being _ble to make predictions of the somlc-boom pressures for the case of a large airplane at high _ltlt_des for waxca i_ n_s u=_- shown theo/etlcaily .... the llft _um- portent of the boom pressure might be relatively large (ref. 5). Experi- mental d_ta under re_listlc flight conditions are urgently needed for 154 .... : .... "".. ' : • • • • • • 4 • • • • • • • • _t_ • r _: ece • • t 40 el I , l, ,• ee l correKation with results of analytical studies and to evaluate the atmospheric propagation losses. An extrapolation of data from fighter airplanes at high altitudes suggests that cruise-flight altitudes in the vicinity of 60,000 to 70,000 feet may be acceptable for the super- sonic transport.

Because of the increased performance capability of some proposed aircraft, it will be possible to operate at supersonic Mach numbers at very low altitudes. The qt'estion has arisen as to the possibility of doing enough damage as a result of the sonic boom to warrant its use as a tactical weapon in a manner illustrated in figure i0. The airplane would be flown on a low-altitude pass over a suitable target area in such a manner as to expose it to damaging pressures in a short interval of time. Sach an operation might have the effect of temporarily inerting some types of enemy activity on the ground over fairly large areas.

Indications of the nature of the pressures obtainable and their effects are given in figure ll. When the airplane is at a relatively high altitude, that is several thousand feet, the pressure signature has the characteristic "N" wave shape as illustrated in the lower sketch. Peak overpressures 2_ up to about lO lb/sq ft are obtainable.

In this pressure range, humans and animals are startled, and damage has occurred to large plate glass windows and to plaster _Blls.

When the airplane is at a relativel_ low altitude, the pressure signature has a shorter period and is m_ch more complex in nature as illustrated in the upper right-hand sketch of figure ll. Peak over- pressures 2__ up to about lO0 lb/sq ft have been obtained for fighter planes operating overhead at a vertical distance of about 150 feet• Up to overpressures _p of lO0 lb/sq ft, no lasting physiological effects were noted for people repeatedly exposed, although they were startled and may have su/'fered some temporary hearing loss. Widespread window damage has occ,Arred, and in some cases buckling of wall and roof panels has occurred. There have also been incidents of malfunction of nonruggedized pressure-sensltive electronic equipmenT,. Measurable verti- cal and horizontal earth motions have been recorded for a wide range of supersonic-flight conditions.

CONCLUDING R_RKS In conclusion, some of the principal noise problems antici_ted for future aircraft types such as the subsonic V/STOL airplane, the supersonic transport, and the special mission aircraft cal_ble of supersonic flight at low altitudes have been dlscus_ed. The main noise sources are noted to be the power plants, the boundary layer, and the shock waves. Engine-noise problems will be of particular concern in • l I ..... , commercial-type operat d' e' 'supersonic transport, par- ticularly during the landing operation. Boundary-layer noise is of importance for aircraft such as the supersonic transport and any special mission aircraft that fly at high dynamic pressures. Special provision will have to be made in the design of the supersonic transport to allow it to operate at sufficient altitudes so as to minimize sonlc-boom dis- turbances on the ground and to avoid damage. The ability of the sonic boom to create some types of structural damage may be used to advantage for special tactical missions.

REFERENCES i. Anon.: Studies of Noise Characteristics of the Boeing 707-102 Jet Airliner and of Large Conventional Propeller-Driven Airliners.

Bolt Beranekand Newman Inc., Oct. 1958.

2. Richards, E. J., Bull, M. K., and Willis, J. L.: Boundary Layer Noise Research in the U.S.A. and Canada; A Critical Review.

USAA Rep. No. 131, Univ. of Southampton, Feb. 1960.

5. Hilton, David A., Mayes, William H., and Hubbard, Harvey H.: Noise Considerations for Manned Reentry Vehicles. NASA TN D-450, 1960.

4. Lina, Lindsay J., Maglierl, Domenic J., and Hubbard, Harvey H.: Supersonic Transports - Noise Aspects With Em@.hasis on Sonic Boom.

2rid Supersonic Transports (Proceedings), S.M.F. Fund Paper No. FF-26, inst. Aero. Sci., Jan. 1960.

9. Walkden, F.: The Shock Pattern of a Wing-Body Combination, Far From the Flight Path. Aero. Quarterly, vol. IX, pt. 2, May 1958, pp. 164-194.

._ ": .:..;... ...

AIRGRAFT TYPES "THRUST, LB 65,000 !20,000 4 0,000 WEIGHT, LB 55,000 350,000 60,000 M 0.6 3D 0.9-2.5 ALTITUDE, FT 40,000 70,000 tO0 Figure 1 PRINCIPAL NOISE SOURCES

POWER 4 4

PLANTS

aOUN_RY 4 4

LAYER

SHOCK wa__s ,/ 4

Fibre 2 ---,,'_m_fftrTAT :. :, ,,, _P .... i3"( • !

m I . Is • ° V/STOL OPERATIONS APPROACH CLIMBOUT /

-r: "--l--

1500 _ - / 1500'- Figure 3

V/STOL TAKE-OFF NOISE

TURBOPROP TURBOFAN ACCEPTABLE ,_ , ACCEPTABLE - \ % ,EW,, I00

v,sToL___

- V/STOL_, \ PNdb _ CURRENT JET \ -CURRENT PR_ TRANSPORT -J TRANSPORT --_ m 80- ,,, t I I !

I I I I 0 2 4 0 2 4 DISTANCE FROM LiFT-OFF, MI.

Figure _ P aal,. ,.6 .a.

.... ,: ". I,.,, .

• ° o. • .... • • o ; ," o o •

"o: : ,'v/_dL' .... '

" L'ANDiNG 'NGISE '

TURBOPROP TURBOFAN I20 - ACCEPTABLE - \ _ ACCEPTABLE --\%%, LEVEL;,

.___L_EV_%

:---_- =_,,_,2Z-- _

I _V/STOL

_ _ _',,_

IO0

k-, b3 PNdb V/$TOI.-P_-__-._.

,._ _ CURRENT JET \ TRANSPORT --_ CURRENT PR ;';'_ "_* _ TRANSPORT --J 8( I ! I l I

0 2 4

0 2 4

DISTANCE FROM TOUCHDOWN, MI.

Flgu._e 5

SUPERSONIC-TRANSPORT ENGINE NOISE

TAKE-OFF LANDING r--CURRENT JET TRANSPORt7 - _ ACCFA_TABLE --h.--,_ / ACCEPTABLE 12o ioo PNdb i 80- I I l I I I j . i I 0 2 4 0 2 4 DISTANCE FROM DIS" ANCE FROM LIFT-OFF, MI. TOUCHDOWN, MI.

1Q':l._z,e 6 _e o • • "59 Q • • ee oo i I$ | | BOUNDARY-LAYER NOISE PRESSURE LEVELS

F

,_L rSTRUOTURAL-I)A'Vc_GE

'"1 | LEVEL ..., _;'::_;:.: • -Tz _2_, ,, • ..,:/: z ;,,., ..

SURFACE-PRESSURE LEVEL.DB Im_ _:::,]" J-- ,".._g., _: 30O i,000 3,O0O DYNAMIC PRESSURE q, LB/SQFT Figuz'e ?

BOUNDARY-LAYER NOISE BOUNDARY-i SURFACE ! _ FREQUENCY PRESSURES !

DISTANGE ALONG FUSELAGE

r_ure 8

• . . • .:.-: : : : :: • . .: :: ...........

SUPERSONIC-TRANSPO_RT CLIMB SCHEDULE eo,ooo[-.

CRUISE..CONDITIONS --.--7---',, 60,000 'i::/

• ,:VEL.-_...... •

zo, ooo L 0 1 2 3 MACH NUMBER Figure 9 SONIC-BOOM DAMAGE DAMAGE F:[l_.re 10

/

4e_ i:)l • • t e • • • • • • ee I | e ee _o _e • • oo t ¢ i " ; J • • '_ e • • SONIC-BOOM EFFECTS GROUND - LO_' ;,LTiTUDE GROUND \_. _ __,. -_"""-.._....j- H'GH ALT,TUDE Z_p, EFFECTS I_/SQ FT STARTLE HUMANS & ANIMALS I - 10 DAMAGE TO LARGE PLATE GLASS SOME PLASTER DAMAGE NO LASTING PHYSIOLOGICAL EFFECTS- WIDESPREAD WINDOW DAMAGE BUGKLING OF WALL & ROOF PANELS IO - IOO SOME MALFUNCTION OF ELECTRONIGS MEASURABLE EARTH MOTIONS 1;'tgu.z'e13.

" '18_r aT 6 PET.CEDING PAGE BLANK IOT FIU D.

i': "." " :.2" "': OF ADV_EED MANNED MILITJ_Y AIRCRAFT By Euclid C. Holleman Flight Research Center

N67-33079

_a_ Melvin Sadoff.

• ._mes Research Center INTRCgYJCTION H Paralleling the large increase in the performance capability of i present airplanes has been the increase in the problems connected with the design and operation of these airplanes. Many methods have _een devised to study these problems, but perhaps no single method of analy- sis has achieved the success and universal acceptance accorded the flight simulator as a design and research tool. The simul_tion of flight is a relatively new art which depends to a large extent on the ingenui_ of the designer of the simulator. Of course, the use of a flight simulator will never replace actual flight. However, because of the increased usefUlness of the s_mulator for airplane design and for the reduction of fl_ght t_e, much more effort is being expended to improve the realism of the flight simulator and to increase its flexibility.

Some of the most use._! simulations have involved the pilot in the control loop. A drswing illustrating a pilot-operated flight simulator is presented in figure I. Illustrated is the flow of lh_ormatlon from the computer to the pilot and back to the computer. The pilot is the key link in closing the control loop.

The l_tional Aeronautics and Space _nir_Istratlon has had consid- erable experience with a wide variety of piloted-flight simulators from simple, inexpensive, flxed-chair types to complex and expensive human centrifuges and variable stability a_d control aiz_lanes. As is indi- cat_ in figure 2, these simulators fall logicall_ into two groups, ground based aud airborne, by virtue of their operating environment.

The fixed-base simulatOr setup is described in figure I. The moving visual environment (fig. 2) refers to a _cme-type simulator or a television-csmera sensor with appropriate projection on a screen in front of the pilot's cockpit. The moving-base simulators provide linear acceleration_ such as the normal-acceleration chair or the human centrlfuge at the Naval Air Develol_ent Center, Johncville_ Pa. Other simulators provide angular acceleration or attitude; an exsmple is the pltch-ro_l chair. _e flight vehicles refer to variable-stability :: :. :. . :.. : . -:: ".:..: : : :'.. - .... .. ..: ".: airplanes (for exmnple, the NASA modified F-IOOC airplane) a variable- stability helicopter, and a variable-stability VTOL (the X-l_). Variable- control-system airplanes have also been tested, as has a variable-control helicopter. The low-dynamlc-pressure airplane refers to reaction- control tests with the F-104, whereas the low-lift-drag-ratio landing tests refer to the simulation of _he X-15 landing with the F-f04.

Some of the typical aircraft design problems that have been studied in varying degrees by using flight simulators are as follows: basic stability and damping requirements, piloting techniques, emergency pro- ccdures, evaluation of displs_s, primary control systems, augmentation systems, landing techniques, and performance and ranging. Much effort has been spent in the areas of stability and control, piloting techniques, and augmentation systems. This backlog of experience has provided con- siderable information on, and insight into, the simulator complexity required for a wide variety of aircraft design problems. The purpose of this paper is to review s_e of the more recent simulator results with emphasis on the airplane-design problem areas. Some of the simulator requirements for V/STOL and a low-altitude attack airplane will also be presented. Areas requiring additional effort are discussed briefly.

Simulators for crew training, however, are not considered in this paper.

SYMBOLS b reference lateral length, ft C Z rolling-moment coefficient _cz _C_ pitchimg-mmaent coefficient _C m Cmq = _-"v/ :....

I • I I • I I I I _ eeI_ i: i, i C reference longitudinal length, ft inertia in roll, slug-ft 2

Ix

inertia in pitch, slug-ft 2

_2

H 2VIx CZp, per sec qoSb - , Der sec 2

%a i x C_a

- qoScZ Cmq , per sec qo sC MSe- Iy Cm5 e, per sec 2 rolling velocity, radlan/sec P pitching velocity, radian/sec q dynamic pressure, lb/sq ft

%

S reference area, sq ft V velocity, ft/sec maximum lateral-control deflection_ radism 8 a maximum longitudlnal-control deflection, re, Jan 5e damping ratio umdsmpel natural frequency, rsdian/sec L'_6 i ' " " .:

i!iii!

el " " ' • le • DI_SSION To be effective, a flight simulator should prampt pilr.t response and comment similar to that obtained during actual flight. Pilot opinion_ then, is the prime measuring device for determining the effectiveness of the simulation. Therefore, program results will be reviewed where pilot- opinion comparlsons between simulator and flight are available.

By using simulators and variable-stability airplanes_ the stability and dsmping requirements for both the longitudinal and the lateral direc- tJ onal modes of airplanes have been studied. Representative results are presented in figure 3 showing areas, obtained in flight with a variable- s:.abiLity airplane, that were considered by the pilots to have satis- factory, unsatisfactoryj unacceptable, and uncontrollable longitudinal characteristics. In order to determine the effectiveness of the simu- lators_ this same range of airplane _ynamies has been investigated by the same pilots by use of a fixed-base and a moving-base simulator (the pitch-roll chair). Figure 4 correlates the pilot-opinion results obtained with the piloted simulator with those obtained in flight. The correlation of both simulators with flight is near perfect until the region of poor airplane dynamic characteristics is reached, where the fixed-base-simulator correlation becomes poor. The moving-base simulator correlates to extr_nely poor dynamics. In fact, _7namic characteristics which were uz, flyable with the fixed-base simulator were controllable with the moving-base simulator and in flight; thus, there is a need for motion stimulus in the case of very poor dynamics. The fixed-base simu- lator, however, was completely satisfactory for a wide range of airplane dynamics including the unstable range of airplane characteristics and gave at least qualitative pilot ratings even in the poorest areas such as high-frequency low damping. Investigations have also been conducted for lateral and directional airplane dynamics and lateral-control coupling, and similar results were obtained.

In addition to the work on conventional aircraft_ considerable ground-based simulator work has been completed recentiy to define control requirements for V/STOL type aircraft. Concurrent flight tests of these V/STOL aircrafb have permitted a preliminary comparison between single-degree-of-fr=edom simulator results ar_ the hovering-control requirements frmm flight tests.

Data obtained during this study are shown in figure 2. It should be noted that the important parameters are control power and damping.

Also shown are the basic control power and damping characteristics ....... "_ _- flight _ .............. Altho-_h .......

are limited, _he single-clegree-of-freedom simulator results would Iz_icate that airplanes C and D fall in the region of satisfactory pitch-control 14"f characteristics, whereas aircraft A and B would be expected to be unsat- isfactory. Similarly, the roll control of airplanes A and C appears to be satisfactory, whereas aircraft B is definitely in an unsatisfactory region. Actual flight evaluations of the pitch and roll controllability of these aircraft are correlated with the pilot opinions from the moving- base simulator in figure 6. Generally, the predicted ratings from the moving-base simulator tests are in fairly good agreement with those from flight; however, they appear, in general, to be optimistic; that is, the simulated airplane was easier to fl_ than the actual airplane. Some of these differences might be attributed to such factors as control-system "deadband" and friction, which were not simulated.

Although no quantitative comparisons are available for fixed- or moving-base simulators and flight evaluations of overall hovering and transition characteristics of V/STOL airplanes_ it is felt that a brief qualitative resum4 of experience to date m_y be of interest. From the pilots' point of view, an analytical six-degree-of-freedom simulation with a moving cockpi_ which provides pitch and roll motion has proven very valuable for pilots' prsctice of expected control problems prior to initial flight tests. The simulator also permitted the pilot to determine piloting techniques for recovery from unusual flight con- dltions. However, because the simulation did not include an adequate presentation of the external visual references that the pilots would have in flight, the pilots observed no direct correspondeuce between hovering height control and transition in the simulator and in flight.

When definite limitations in the simulation have been noted on the piloted-flight simulator such as Just described, it has been helpful for the pilot in evaluating a new configuration to fly the simulation of an airplane with which he has had recent flight experience. This procedure serves to orient or calibrate the pilot to the limitation of the simulation so that he can evaluate objectively the relative diffi- culty of the new airplane control task.

Recent pilot evaluations of fixed-cockpit simulators, which provide slx-degree-of-freedom simulated external visual environment, have indi- cated that this type of simulator is admirably suited to the V/STOL simulation problem, particularly for accurately evaluating the hovering and transition characteristics of the airplane. The addition of three- axis angular motion ms_ be desirable but, perhaps, is not essential for this problem.

Another design problem in which the simulator has been used is fc _ checking the pilot's .wresentation. Tests have been made with an airplaz, c, a moving-base s:mulator, and a flxed-base simulator to compare the pilot's performance while tracking with an Inside-out and an outslde-ln target displ_. The performance of the pilots was very poor with the outside- in dlspl_ for both the flight and moving-base simulator, whereas the

i i

i! !i • .....

performance with the inslde-out displ_ was acceptable. These results did not correlate, however, and thus some basic deficiency in the pres- entation or motion stimulus was indicated. With the fixed-base simu- lator the pilot's performance with either of the displays was compa- rable and showed the absence of motlon-stlmulus effects. From these tests, it was concluded that a fixed-base simulator should not be used for the evaluation of tracking displsys and that the results frum movlng-base simulators should be extrapolated to flight only with reservation.

The flxed-base simulator has also been used during the design of airplane instrument displ_ys. Early in the piloted simulator program of the X-19 airplane a scanning problem was noted by the pilots and, as a result, a rearrangement and a consolidation of the panel instru- ments was made. Tests :_th a movlng-base simulator (centrifuge) con- firmed the improvement afforded by these changes. No new presentation deficiencies have arisen during current flight tests.

A requirement has been indicated for several types of manned mili- tsry airplanes. One example is the low-altitude attack airplane. This airplane is not too unlike conventional airplanes and, as with any new development program, design and operational problems are expected. Some of these problems are longitudinal- and lateral-control sensitivity, response to turbulence, and control and aerodynamic coupling.

Previous programs have indicated that these problems can be resolved by using _ flxed-base simulator with one exception, the piloting problem encountered with a hlgh-performance airplane in turbulent air. Recent tests have shown that both controllability and pilot fatigue are impor- tant under these Conditions. A moving-base simulator which duplicates the normal acceleration of the airplane will be required for this prob- lem. Figure 7 illustrates such a simulator, the NAA g-seat. This type of simulator is a relatlvel_ inexpensive piece of hardware and could, it appears, Justify its cost for the investigation of this one problem.

The inclusion of pitch and bank angle of this simulator would add realism but would probably not be required.

Thus far, speclflc-design problem areas that have been investigated on siemlators sad in flight have been d/scussed. In order to illustrate further the importance of the piloted-fllght simulator, a design program that probabl_ woukl not have been possible without the piloted-flight simulator - the X-19 research airplane program - is considered briefly.

Flight simulators dictate_ ma_ important design changes to the airplane, but perhaps their most important contribution was to emphasize the need for a complete simulation. The difficulty of the control task during certai_ parts of the flight envelope showed the need for a movie-base simulation program to investigate the capabilities of the pilot while subjected to the accelerations expected of the airplane. Consequently, • • e • • a program was conducted by utilizing the human centrifuge to impose tile expected acceleration on the pilot while piloting the simulated X-19 mission. The mechanization of the centrifuge for this program is shown in figure 8. During this simulator program it was deteNnlned that, even at the highest anceleration expected, there was little deteriorat_ ,n in the pilot's performance; thus, if the accelerations are below t_.. physi- ological lirlit of the pilot, his performance will be unaffected.

Exposure to the expected accelerations increased the pilot's confide.ce in his ability to cope with the problems of actual flight. Exper/, r:ce from several centrifuge programs has shown that, to determine the tol- erance limit to acceleration, a centrifuge iz necessary; however, foe the investigation of airplane control problems with the centrifuge, serious problems have been noted because of spurious motion cues.

At present, a complete six-degree-of-freedom fixeu-base X-15 simu- lat._r; including the control-system hardware, an airplane-llke cockpit with _nctional pilot's controls, and actual electronic components of the stability-augmentatlon system, is being used for flight planning, pilots' practice for flight, and for verification of airplane flight behavior after flight. The pilots have enthusiastically endorsed the use of the flxed-base piloted-flight simulator for becoming acquainted w_th the piloting task before actual flight. Perhaps the most signif- icant contribution of The X-15 simulator program w_ ii be correlation of the data from flights, moving base simulators, and fixed-base simu- lators for defining the simulator requirements for the design of future manned military and research airplanes.

Some results of simulator studies which have provided information concerning the type of simulator best suited for different investigations have been described. A few areas where continued effort would result in large dividends will now be discussed.

While investigating the steep-glide approach to a landing, by using the piloted-flight simulator, it was necessary to resort to actual flight with a test airplane because of the lack of realism of the simulator with a conventional presentation. In this area the simulation of the airplane flight environment by televised projection would be admirably suited.

While a moving visual environment is being considered, another area requiring continued study is the blending of visual and motion stimuli on the simulator. An exsmple of the effective use of this blending is the DC-8 simulator, in which the initial angular-acceleratlon motion is simulated az_ the motion effect is continued by the visual environment.

Pilots report that this simulation of flight is very realistic.

In order to simulate adequately problems of long duration that cover s wide range of operating conditions (navigation, for example), greater accuracy of the analog computer is required. Digital-computer elements eo eee • (lee ¢ ee • 4, • e I_0 ' ' • • el • el a I • • a , o o l ee eli I and converters are available; however, the present cost of this equipment for most piloted-flight-slmulator applications may be prohibitive.

CONCLUDING REMARKS The fixed-base simulator wlth adequate presentation and controls is satisfactory for the investigation of a wide range of airplane prob- lems; however, there are areas where realism can be er_anced by suit- able motion sm_ visual-environment stimuli. Concerted effort, to increm_ the usefulness and realism of the simulation will yield large dividends in the form of reduced costs of design and flight test of manned air- planes. Finally, caution should be exercised in mechanizing the piloted- flight simulator to avoid unnecessary complexity and costs which would actually retard the develol_ent of tLe airplane.

!

L

iDl . Q Q •

i'• .. "°" '

• • • _ee I; ee ,Qe • BIBLIOGRAPHY Brown, B. Porter: Ground Simulator Studlez of the Effects of Valve Friction, Stick Friction, Flexibility, _ Backlash on Power Control System Quality. NACA Rep. 1548, 19_8. (Supersedes NACA TN 3996.)

Brown, B. Porter, and Johnson, Harold I. : Moving-Cockplt Simulator Investlgat[on of the Minimum Tolerable Longitudlnal Maneuvering Stsbillty. NASA TN D-26, 1959.

Brown, B. Porter, Johnson, Harold I., and Mungall, Robert G. : Simulator Motion Effects on a Pilot's Ability To Perform a Precise Longitudinal Flying Task. NASA TN D-367, 1960.

Creer, Brent Y., Heinle, Donovan R., and Wingrove, Rodney C. : Study of Stability and Control Characteristics of Atmosphere-Entry Type Aircraft Through Use of Piloted Flight Simulators. Paper No. 59-129, Inst. Aero.

SCl., Oct. 9-7, 1999- Creer, Brent Y., Stewart, John D., Merrick, Robert B., and Drinkwater, Fred J., III: A Pilot's Opinion Study of Lateral Control Require- ments for Fighter-Type Aircraft. NASA MEMO 1-29-59A, 1999.

Douv_llier, Joseph G., Jr., Foster, John V., and Drinkwater, Fred J., III: An Airborne Simulator Investigation of the Accuracy of an Optical Track Command Missile Guidance System. NACA RM A96G24, 1996.

Douvillier, Joseph G., Jr., Turner, Howard L. McLean, John D., and Heinle, Donovan R. : Effects of Flight Simulator Motion on Pilots' Performance of Tracking Tssks. NASA TN D-155, 1960.

Eggleston, John M., Baron, Sheldon, and Cheathsm, Donald C. : Fixed-Base Simulation Study of a Pilot's Ability To Control a Winged-Satelllte Vehicle During High-Dra_ Variable-Lift Entries. NASA TN D-228, 1960.

Foster, John V., Fulcher, Elmer C., and Heinle, Donovan R. : An Air- Borne Target Simulator for Use With Scope-Presentation _ Fire- Control Systems. NACA RM A57C19, 1957.

Hardy, James D., 8_. Clark, Carl C. : The Development of Dynamic Flight Simulation. Aero/Space Eng., vol. 18, no. 6, June 1999, pp. _8-92.

Holleman, Euclid C., Armstrong, Nell A., and Andrews, William H. : Utilization of the Pilot in the I_mnch and Injection of a Multistage Orbital Vehicle. Paper No. 60-16, Inst. Aero. Sci., Jan. 25-27, 1960.

• • . ° • . ..

152 .......

Holleman, Euclid C., and BosLaugh, David L. : A Simulator Investigation of Factors Affecting the Design an_ Utilization of a Stick Pusher for the Prevention of Airplane Pitch-Up. NACA RM H57J30, 1958.

Hollaman, Euclid C., and Stillwell, Wendell H. : Simulator Investigation of Co, and Reaction Controls. NACA RM H_gD22, 1958.

James, Harry A., Wingrove, Rodn_y C., _.olzha_ser, Curt A., and Drinkwater, Fred J., IlI: Wind-_irmel ar_ Piloted Flight Simulator Investigation of a Deflected-Slipstream VTOL #.trplane, the Ryan VZ-3RY. NASA TN D-89, 1959.

Matthews, Howard F., and Merrick, Robert B. : A Simulator Study of Some Longitudinal Stability and Control Problems of a Piloted Aircraft in Flights to Extreme Altitude and High Speed. NACA RM A56F07, 1956.

McFadden, Norman M., Paulij Frank A., Eeinle, Donovan F. : A Flight Study of Longitudlnsl-Control-System Dynsmlc Characteristics by the Use of a Variable-Control-System Airpl_ne. NACA RM A97LlO, 1958° McNelll, Walter E., and Creer, Brent Y. : A Su_nary of Results Obtained Durlr,g Flight Simulation of Several Aircraft Prototypes With Variable- Stability Airplanes. NACA PJ4 A56C08, 19%.

Ra_hert, George A., Jr., C_eer, Brent Y., and Douvll!ler, Joseph G., Jr. : Use of Flight Simulators for Pilot-Control Problems. NASA _M0 3-6-59Aj 1959.

S_ioff, Melvin: The Effects of Longitudinal Control-System Dynamics on Pilot 0pinion and Response Characteristics As Determined From Flight Tests and From Ground Simulator Studies. NASA MEMO 10-I-58A, 1958.

Stillwell, Wendell H., and Drake, Hubert M. : Simulator Studies of Jet Reaction Controls for Use at High Altitu_e. NACA RM _SG18a, 1958.

Well, Joseph, and D_, Richard E. : An Aual_g Study of the Relative Importance of Various Factors Affecting Roll Coupling. NACA RM R_6A06, 1956.

White, Maurice D., a_l Drinkvater, Fred J., III. A Compscison of Carrier Approach Speeds As Determined From Fllght Tests and From Pilot-Operated Simulator Studies. NADA RM A57D30, 1957.

1_ __:i.I

FIXED-BASE SIMULATOR

PRESENTA]., )N SIGNALS CONTROL b;$NALS E-5636 Figure i

TYPESOF SIMULATORS

GROUN_ BASED 1

I ]

FIXED, _ ] i !

L

I I [ , ATTITUDE

_°_ I I _vllG VlSUll I I I I "

, _T AI_LEIATION I ACTUAL FLIGHT I

I

LOW LID

I1

Figure 2 15_ l 11! ! | $ I,_O el Oe • % • e _ • • • B • • I i | I e m t _ • • • i 1 J e • i _ • lie • FLIGHT EVALUATION OF AIRPLANE DYNAMICS I S FAC T ORY 301 i _a._r_ ,.-- , 2c o° PER SEC 2 I I q2 4 6 2_(_ n, PER SEC Figure 3 CORRELATION OF PILOT OPINION LONGITUDINAL DYNAMICS LINE OF PERFECT CO R R E L AT ION'_ / // /

%.-

FLIGHT FIXED BASE MOVING BASE (PITCH-ROLL CHAIR.} Figure 4 de ,. : . :'_;;._ , • , , ;-:,," .,' : V/STOL HOVERING CONTROL REQUIREMENTS PITCH ROLL -2.5 -2.C Mq, -1.5 • T PER SEC l'b_-'# "IY' PER SEC-I.C ATI SFACTORY X9 ,-4 A C I 0 .5 1.0 1.5 2.0 2.5 0 I 2 3 4 5 M_)e_e, L_a_a, PER SEC 2 PER SEC 2 Fi£ure9

C, ORRELATION OF PILOT OPINION

V/STOL CONTROL ROLL PITCH d /

°%,,/

/-%

B /

%/I

FLIGHT i n | MOVI NG BASE MOVING BASE (PiTCH-ROLL CHAIR) (PITCH-ROLL CHAIR) Yi_e6

NAA g- SEAT

Figure 7 E-5665

CENTRIFUGE DYNAMIC SIMULATION

ACCELERATION O,JTPdTS i DISPLAY OUTPUTS PILOT CONTROL MOTIONS E-56_7 Fi_e8 i o • • • • • . • t_e e _e _e _st_ VTOL aircraft: Helicopters will continue to be the best VTOL for missions where long hovering time is required and where low speed and short range are acceptable. Compound helicopters are expected to give a small increase in cruising speed and range. The propeller-drlven tilt-wlng and flap configuration is considered to be one of the most promising VTOL types, particularly for use on transport missions where long range is required and where higher speed is advantageous. Although a great deal of research and development will be required before optimum operational VIOL aircraft will be obtained, progress in thls field has advanced to the point where operationally useful machines of the most promising types can be designed and built. A great need at the present time is for experience with VTOL aircraft to demonstrate their potenti- alities and to define more clearly the service requirements in the va_'ious areas.

Variable-sweep multimlssion aircraft: Recent research on variable- sweep wings has led to the development of configurations having accept- able stability and control characteristics over a large wing sweep angle variation without the previously required wing translation. This development opens a new potential for improving the compatibility of the configurations needed for optimum performance at supersonic and subsonic flight conditions, and thus offers greatly improved aircraft versatility. It appears, for example, that one properly designed varlable-sweep airplane will be able to accomplish a number of impor- tant missions such as transoceanic ferry, extended subsonl_- patrol or loiter, long-range hlgh-altltude supersonic attack, and long-range low- level supersonic attack. Efficient STOL performance also can be pro- vialed in the same aircraft If desired, in order to permit operation out of very small fields. It is estimated that the weight and size of such a multimission aircraft will not appreciably exceed the weight and size of any of the specialized slngle-purpose aircraft which it can replace• Supersonic cruise aircraft: Supersonic cruise efficiency comparable to that of present subsonic Jet transports has been shown to be attain- able within the present state of the art. Recent research indicates further potential gains by reduction of turbulent skin friction through boundary-layer injection mud by improvements in drag due to lift. For commercial supersonic transports the major aerod_c problems are found in the off-design areas, such as take-.off and landing, transonic acceleration at high altitude, and subsonic cruise efficiency.

Air-breathing propulsion systems: The current status of propulsion system components designe_ for Mach 3 flight has been reviewed, and it has been shown that reascaably high levels of on-design performance can be obtained. Off-design performance improvements are necessa_j and will • . • • • • • require a great deal of individual tailoring. The turbofan engine appears to offer advantages over the turbojet in the areas of take-off noise, off-design upeci_Ic fuel consumption, higher augmentation ratios at transonic speeds, and lower temperatures during supersonic cruise.

Some disadvantages may be a greater frontal area for a given thrust and larger inlet stud ductlng weights than the turbojet. In spite of these disadvantages, the desirable characteristics appear to make further development of the turbofan highly warranted.

The X-15 flight research program: The research objectives of the X-15 flight program were formulated to provide information over a wide range of conditions pertinent to the development of advanced military aircraft. The flight results obtained thus far indicate reasonable agreement with wind-tunnel predictions regarding aerodynamic forces and heating up to a M_ch number of 5. Current plans are to extend the data on the X-l_ to these and other aerodynamic, structural, and systems problems in the speed range between Mach numbers of 5 and 6.

Hypersonic-cruise vehicles: Aerodynsmic lift-dr8_ ratios have been obtained on both hydrocarbon-fuel and hydrogen-fuel vehicles which are high enough to provide desirable cruise ranges. The thermodynsmics of the ramjet units yield values of propulsive efficiency which, when coupled with the aerodynamic efficiency, indicate desirable range- payload possibilities. The aerodynamic heating of the vehicles is low enough that it is within the present construction capability of the aviation industry.

Structures and materials: Structural design and structural mate- rials have an important bearing on the integrity and performance of military aircraft. Weight-strength considerations are useful in the selection of materials; however, other factors such as tear resistance are becoming increasingly important, particularly for transport-type aircraft. Various types of structural construction are of interest for high-density heat-resistant materials. These include honeycumb- sandwich, open-face sandwich, and skin-stringer types. Among these the honeyccmb-sandwleh construction generally yields the lightest structure.

For wing-type structures the most efficient honeycomb-sandwich construc- tion yields a deeper wing than either the most efficient skin-stringer or open-face sandwich. Other factors that are important in aircraft structures include strength under nonuniform temperatures, sonic fatigue, and panel flutter• Creep, on the other hand, will in all probability not be a major structural problem.

Noise considerations: The m_in noise sources anticipated for i_ture aircraft types such as the subsonic V/STOL airplane, the supersonic trade,oft, and the siJeclal mission low-altitude supersonic aircraft, are noted to be the power plants, the boundaz_ layer, and the shock waves.

Engine noise problems will be of particular concern in cummercial t_pe !

.......... 1D'J _ _o oe I , ' t • i • , I i !

!i' :,,' ' ,',,,, ..... ''

• I i I I I I i 1 I II operations of V/STOL and the supersonic transport with respect to adverse reactions in communities near airports, particularly dur|n_ the landing operation. Boundary-layer noise is of importance for air- craft such as the supersonic transport and any special mission aii'craft that fly at high dynamic pressures. Special provision will have to be made in the design of Lhe supersonic transport to allow it to operate in such a way as to minimize sonic boom disturbances on the ground and to avoid damage to ground structures. The ability of the sonic boom to create structural damage may be used to advantage for some special tactJ cal mlssions.

Simulation requirements: The flight simulator has been un_versally accepted as an effective tool for manned aircraft design and operational research. The fixed-base simulator with adequate presentation and con- trols has been satisfactorily employed for investigstion of a wide range of problems. Furthermore, its value may be enhanced by the inclusion of certain motion and visual-environment stimulij such studies, for example, would be useful in the evaluation of hovering and transition character- istics of V/STOL aircraft or control and response of the low-level attack airplsr_e. Televised projection techniques and the blending of such visual stimuli with motions are expected to improve the simLllation of flight environment for low-level flight and landing approaches. However, caution must be exercised to avoid prohibitive complexity and cost in simulators, even though the technology for provision of more complete simulations is available.

National Aeronautics and Space Aaministration, Washington 25, D.C., August 26, 1960.

NASA- Uw_ ,i,m. v,. L-12-10 - 1215, 1279, H-182 - 185