Comparison of model and flight test data for an augmented jet flap STOL research aircraft

General Disclaimer One or more of the Following Statements may affect this Document This document has been reproduced from the best copy furnished by the organizational source. It is being released in the interest of making available as much information as possible.

This document may contain data, which exceeds the sheet parameters. It was furnished in this condition by the organizational source and is the best copy available.

This document may contain tone-on-tone or color graphs, charts and/or pictures, which have been reproduced in black and white.

This document is paginated as submitted by the original source.

Portions of this document are not fully legible due to the historical nature of some of the material. However, it is the best reproduction available from the original submission.

Produced by the NASA Center for Aerospace Information (CASI)

NASA TM X 62 ,491 NASA TECHNICAL MEMORANDUM N76-10093 COPPARISON OF MODEL AND (NASA-TM-X-624 0 1) FLIGHT TEST DATA FOR AN AUGMENTED JET FLAP a 45 p HC $3.75 STOL RESEARCH A111CRAF1 (NASA) Ur ► clas CSCL 01C X G3/05 38437 Q H Q COMPARISON OF MODEL AND FLIGHT TEST DATA FOR AN f AUGMENTED JET FLAP STOL RESEARCH AIRCRAFT r+ W. L. Cook and D. C. Whatley Ames Research Center Moffett Field, Calif. 94035 and de Havilland Aircraft of Canada Ltd.

Downsview, Gntario, Canada June 1975 .ti w Rv^!

Sp BR^^^/^Tr ^L:

I Report 2 Government ArteMion No Recipient's Catalog No No. 3 TM X-62,491 4. Title and Subtitle 6 Report Date COMPARISON OF MODEL AND FLIGHT TEST DATA FOR AN Performing Organization Cods, AUGMENTED JET FLAP STOL RESEARCH AIRCRAFT Performing Organization Report No 7. Authorls) 8 A-6303 W. L. Cook and D. C.

Whittley 10 Worts Unit No 9 PerformingOrganintion Name and Address 514-50-01-01 NASA Ames Research Center 11 contract or Grant No Moffett Field, Calif. 94035 and de Havilland Aircraft of Canada Ltd.

Period Canada Downsview, Ontario 13 Type of Report and Covered 12. Sponsor;ng Agancv Name and Address Technical Memorandum Natioual Aeronautics and Space Admii,istration 14 Sponsoriry Agency Code Washington, D. C.

20546 pplementar y Notes 16. su 16 Abstract .

Aerodynamic design data for the Augmented Jet Flap STOL Research Aircraft or commonly known as the Augmentor-Wing Jet-STOI, Research Aircraft was based on results of tests carried out on a large scale research model in the NASA Ames 40- by 80-Foot Wind Tunnel. Since the model differs in some respects from the aircraft, precise correlation between tunnel and flight test is not expected, however the paper delineates the major areas of confidence derived from the wind tunnel tests and shows that, for the most part, tunnel results compare favourably with flight experience. In some areas the model tests were: known to be non-representative so that a degree of uncertainty remained: these areas of greater uncertainty are identified and again discussed in the light subsequent flight tests.

18 Distribution Statement 17 Key Words (Suggested by Authorlsl 1 Unlimited STOL Augmented jet flap Aerodynamics 'TAR Category — 02,05, 07,0 Price' No. of Page 22 Security Classif. (of this pate) 21 20.

19 Sacunty Classif, (of this report) Unclassified Unclassified 22151 'For sale by the National Technical Information Service, Springfield, Virginia

NOTATION A augmentation ratio • Cp drag coefficient, dra g XB blowing momentum coefficient, C qs lift coefficient, lift C L rolling moment coefficient C XV C T thrust coefficient vectored hot thrust qs engine high speed N N V velocity engine fan air blowing thrust

x

1{ 1 engine vectored hot thrust C wing chord length h height above gruund q free stream dynamic pressure CL wing angle of attack mass density c` ambient air P 6 A aileron deflection 6 F trailing edge flap deflection angle elevator deflection angle 6 e 6 TH vectored hot thrust nozzle deflection angle 6 control wheel angle w n nozzle efficiency • roll angle m roll acceleration, rad/sect

Subscripts A free air or aileron slideslip angle B F trailing edge flaps isentropic I G in ground effect iv

COMPARISON OF MODEL AND FLIGHT TEST DATA FOR AN AUGMENTED JET FLAP STOL RESEARCH AIRCRAFT by W. L. Cook Chief, Research Aircraft Projects Office NASA, Ames Research Center Moffett Field, California 94035 and D. C. Whittley Manager, Research & Augmentor Technology De Havilland Aircraft of Canada Limited, Downsview, Ontario, Canada SUMMARY Aerodynamic design data for the Augmented Jet Flap STOL Research Aircraft • or commonly known as the Augmentor-Wing Jet-STOL Research Aircraft was based on results of tests carried out on a large scale research model in the NASA Ames 40- by 80-Foot Wind Tunnel. Since the model differs in some respects from the aircraft, precise correlation between tunnel and flight test is not expected, however the paper delineates the major areas of confidence derived from the wind tunnel tests and Shows that, for the most part, tunnel results compare favourably w + th flight experience. In some areas the model tests were known to be non-representative so that a degree of uncertainty remained: these areas of greater uncertainty are identified and again discussed in the light subsequent flight tests.

INTRODUCTION Research programs relating to boundary layer control for high lift were quite common in the time period 1955 - 1965. NASA was particularly inter- as applied to a possible sTOL transport and modified ested in the con cept two aircraft for research: the "Lockheed BLC Hercules" and the "Boeing 707-80". Generally speaking, it was found that the maximum lift which could be achieved by flap blowing (with just sufficient thrusv to maintain attached flow) was only slightly higher than that for a good double slotted mechanical flap. Also, engine thrust loss due to compressor bleed and to the ducting, degraded takeoff performance of the projected "STOL airplane". In the light of this experience, interest was revived in jet-flap type of blowing concepts which utilize a significant proportion of total thrust in the flap and oper- ate in the regime of supercirculation to yield much higher values of lift coefficient. It was in this context that a mutual interest was established between scientists in Canada and the U.S.A. in the augmentor flap (fig. 1) in A -6303

1963/64. It was foreseen that the penalty of duct loss could be off -set by a thrust increase generated by the augmentor flap while, at the same time, engine thrust loss c^uld be avoided by using bypass or fan air to blow the flap in the place of compressor bleed air. In addition, the two element f'.ap which completely shrouded the ,jet would ensure a very positive and substantial degree of thrust vectoring which is so essential for a steep gradient approach with lending flap setting.

These were the bread motivations but it was n p r_essnry to demonstrate many aspects of aerodynamic performance and stability before feasibility of the concept for a flight test aircraft could be established. In both countries, it was concluded thet tests at large scale were essential because ejector sys- tems were known to be sensitive to small changes in geometry and surface Irregularities. Thus, in 1964, agreement was reached between NASA, Ames Research Center, the Canadian Defence Research Board and de Havilland to design and build a large scale model based on the Augmentor-Wing concept for test in the Ames 40- by 80-Foot wind Tunnel (fig. 2). The successful outcome of these tests led to consideration of a research aircraft based on the de Havilland Buffalo airframe and using the Rolls-Royce Spey turbo-fan engine. Accordingly, in 1968, the existing large scale nodel was fitted with nacelles and pylons each containing a General Electric J-85 engine with vgctored thrust while the the tailplane height and size were adjusted to be more representative of the Buffalo aircraft. Following these modifications the model underwent a se ies of intensive tests in the Ames 40- by 80-Foot Wind Tunnel, with the model the tunnel mounted both at the center of the tunnel and in close proximity to floor to simulate ground reported in reference 1. Even though the layout of the research aircraft (fig. 3) was not established In detail until mid-1970, it was t hese model tests which formed the main source of design data.

Although many differences exist between the model and the airplane, gener- ally it was agreed that the two were similar in all major respects so tb.,.r cost of additional wind tunnel tests could be avoided. It is the purpo this paper to show that reasonably good correlation has been found betwe tunnel, simulator and flight tests in most respects and to discuss certain areas where some real doubt existed about the predicted flight characteristics because of gaps in the design data base.

GEOMETRY AND CONFIGURATION COMPARISONS General arrangement drawings of the model and the research aircraft are shown in figures 4 and 5 respectively. Some pertinent geometric data is shown in the table 1. The model is approximately a 0.55 scale of the aircraft.

The model had a wing area of 21.4 meters squared whereas the research air- craft has a wing area of 80.4 meters squared. The major differences between the model and the aircraft that would have an effect on the aerodynamics are the wing aspect ratio, 8.45 (model) compared to 7.2 (aircraft), the augmenr_or flap to wing chord ratio, 0.23 (model) compared to 0.24 (aircraft), the wing A-6303

dihedral, 0-degrees (model) compared to 5-degrees (aircraft), and the model ailerons were tapered whereas the aircraft ailerons were constant chord length.

The research model is powered by a turbo-conpressor unit consisting of one General Electric J-85 engine, used as it gas generator to drive two modified Rolls-Royce Viper compressors. In addition, each nacelle contains one General Electric J-85 turbo-jet with diverter valve for thrust vectoring. Thus blow'., thrust and nacelle thrust could be varied independently in the wind tunnel.

The research aircraft is powered by two Rolls-Royce Spey turbo-fan engines modified to collect the fan air for wing blowing and fitted with Pegasus type nozzles to vector the hot thrust. The correspondence of the "cold" blowing, thrust to the "hot" vectored thrust for the modified Spey engine is shown in figure 6.

BAST ..ORRELATION As detailed in the previous se.:tion and table 1, some differences exist in geometry and configuration between the wind tunnel model (fig. 4) and the research aircraft (fig. 5). The performance and stability of the research aircraft were estimated prior to firzt flight using results from the wind tunnel as the primary source of design data with due allowance being made for these differences. Correlation is based on a comparison of the flight data with these prior estimates but, wherever possible, reference is made directly to the tunnel test results to illustrate the predicted trends and assess the degree of correlation.

Th- landing configuration is the wore critical one from an aerodynamic point or view in nearly every respect: likewise, it is more demanding from an operational point of view. Therefore, in the present paper, consideration has been given primarily to configurations with flap angles in the range 6 F = 650/750.

COMPARISONS OF MODEL AND FLIGHT DATA Eight aspects of aerodynamics and design have been selected for discus- sion as follows: Duct loss and thrust augmentation Performance • - Stalling characteristics Ground effect Longitudinal stability Late.al stability Roll control power Flight simulation

A -6303

Duct Loes and Thrust Augmentation The pressure loss associated with the ducting of an internally bloom flap presents a source of inefficiency in this class of powered lift concept. How- ever, one objective of the DHC/NASA program was to demonstrate that a signifi- cant quantity of air flow could be ducted to the flap without undue loss and that, in any event, ar. increase in thrust could be achieved in the alAgmentor flap which would more than off-set any thrust degradation on this account. A further objective was to c monstrate that the air-flow could be accommodated in the wing using a relatively simple ducting/nozzle combination located behind the wing spar thus leaving the main wing box essentially unaffected.

This implied that the wing nozzles would be "end fed" from spanwise circular ducts rather than by fish-tail type ducting as in the case of the British jet- flap research aircraft. In general, tt has been our aim to maintain the duct Mach numer at less than 0.30 for choked conditions at the nozzle and then it becomes possible to maintain a fairly constant pressure along the duct (and at the nozzle) for all higher or lower values of pressure ratio.

Uuet Loss.— The duct layout in the model is different to that in the research Aircraft but the general constraints and objectives are the same, namely, to deliver about 40 percent of the engine thrust to blow the wing with a minimum of loss.

Figure 7 illustrates the loss in total pressure from the final compressor stage to the ducts which contain the wing blowing slot of the large scale model. The pressure loss is in the order of 14 percent which corresprnds to a thrust loss of 9 1/2 percent. An extensively modified Viper engine +as used as a compressor unit which, on account of certain design compromiser led to a significant pressure loss between the final compressor stage and ie com- pressor plenum. Therefore a more realistic appraisal of the penalt associated with ducting can be obtained on the basis of loss between the come 2ssor plenum and the augmentor nozzle duct. On this basis, after remo , _,ig 4 percent due to the collector system, the results show a thrust loss of about 5 1/2 per- cent with more detailed information shown in figure 7.

A schematic of the ducting in the research aircraft is given in figure 8: also shown is the total pressure and thrust loss between the engine compres- sor and various points in the ducting. On the average, duct loss is less than 10 percent and the corresponding thrust loss less than 5 percent - again with more detailed information being given in the figure. Thus, achievement in this regard on the research aircraft. is at least equal to that on the model in spite of the additional requirement to provide a cross-ducting capability.

It is difficult to obtain an accurate measurement Thrust Au^ntaticn. — of static thrust augmentation on the research aircraft largely because of the predominant primary hot thrust of the Rolls-Royce Spey engine. However, a fully representative 7/10 scale model of the ducting system and flap was tested by Boeing during the design phase. This model, with a span of 2.4 meters, incorporated the 'twin-nozzle' arrangement as shown in figure 1. Results have been compared to half scale model of 0.38 meters span tested in the de Havilland laboratory (fig. 9).

4 A-6303

note that nozzle loss d •s not appear directly as part It is important to of ''ie system performance because thrust augmentation is defined to terms of a net gain relative to the isentropic thrust supplied to the augmentor system (where isentropic thrust is based on duct supply pressure and mass flow). The as follows: Net augmentation a Gross net augmentation of thrust is defined ' N A where Agross augmentation X Nuzzle efficiency, i.e., net - Agross. r Measured augmentor thrust i Measured nozzle thrust.

correction to a coi7var&ble value of L/t ratio.(augmentor length/ After nozzle thickness) the de Havilland model for flap angle dF - 50 0 , gave Agross 1.45 and Anet ' 1.37 (corresponding to a nozzle efficiency of 0.94). By com- parison the Boeing 7/10 scale flap model of the research aircraft gave AgrQas ' 1.27 (corresponding to a nozzle efficiency of 0.92). ReRu is 1.38 and Anet of tests for a range of flap angle are showr in fig. 9 in which the Boeing reference 2.

test data are taken p rom Both models demonstrated the capability of increasing augmentation by 4 or 5 points when operating either the -upper or lower nozzle singly.

The larger scale and larger span model was unable to generate a static thrust augmentation equal to the component model in the laboratory. This experience has been fairly general throughout, whenever large scale augmentor conducted. In part, this can be attributed tn inter- exper'ments have been feren, " e caused by structural members located inside the augrentor passage and due to the sensitivity of a slot nozzle ejector to such disturbances.

Performance Correlation has been studied for two configurations: approach and landing flap, where the ability to perform a steep descent gradient is important and takeoff flap, where climb gradient is important especially in the case of engine failure.

The appropriate aerodynamic coefficients are defined as follows: This coefficient includes the thrust of the jet flap but excludes CLnet any lift component of hot jet reaction.

This coefficient is a measure of "drag minus thrust". It includes CDnet jet flap thrust but excludes thrust component of hot jet and momentum drag of the J-85 engines.

CJ The blowing coefficient is based on the isentropic thrust avail- able in the wing ducts and is obtained from a knowledge of duct pressure and mass flow. It includes the blc thrust associated with the fuselage and ailerons which represents about 10 percent of the total.

Total lift coefficient, including lift component of hot thrust.

CLT a two main differences between the wind tunnel model and the air- There craft which affect performance; aspect ratio 8.45/7.2 and flap-chord ratio .33/.24 (model to aircraft respectively). Data from the model tests have been corrected for these differences according to formulae given in reference 3 aad

4 A-6303

applied to a configuration which was close to trim. The magnitude of the correction is shown in figure 10 for landing fl ^ and found to be quite sig- nificant.

Wind tunnel performance data are obtalne: by running polars at constant speed and power (i.e. constant blowing coefficient). In flight, as the aircraft executes a polar at constant power, the airspeed decreases as angle of attack increases and therefore it is more difficult to obtain a set of parametric flight data. An analysis method has been used wherebv and for a CLnet Cpnet given flight polar are plotted against angle of attack; values are interpolated at constant values of a w (for each polar) and then plotted against blowing coefficient, C J .

It then becomes possible to construct lift and drag polars with Q as parameter.

variation of versus Zpp roach and landing flap .— , The for aw CLnet CJI 7.5 0 and for aw - 17.5 0 are shown In figure 11. It was found that lift coefficient falls below prediction for a given aw whereas the variation with follows the predicted trend quite well.

blowing coefficient (at constant aw ) It is of interest to note that "single engine" points fall in line with the two-engine data.

Tte drag polar is of special interest because the vs CL relationship CD provides a measure of descent gradient capability. Flight data show good agreement with model tests a low values of CJ (see fig. 12) but Exhibit very little thrust recovery as C J increases. Thus the flight results depart more and more from the model tests as the value of CJ increases. Generally, it is thrust recovery will be realize: provided the jet sheet is ben, thought toat backward by the free stream before !t breaks up (since after break-up it cai.

no longer sustain a pressure differential). It is suggested that the constraint of the tunnel floor may have caused the jet sheet to bend backward sooner (for these large flap angles) and thereby some thrust recovery is achieved in the tunnel whereas, in flight,the jet sheet breaks up before being deflected streamwise. In any event the trend of the flight result is favourable, in that it permits the achievement of a steeper descent gradient.

Takeoff flap configuration.— The variation of lift coefficient with wing - 30 0 , The correla- blowing thrust is shown in figure 13 t.,- takeoff flap, 6 F tion is reasonably good and somewhat better than for 6 F - 65''. The correspond- ing drag polars in figure 14 compare well with the estimate and indicate a similar degree of thrust recovery to that found in the tunnel. Accordingly, it was found that predictions of climb performan^e for takeoff flap setting agreed well with flight data and, in particular, for the single engine case.

CLmax and Stall Characteristics Correlation of the stall is presen'. ed in figure 15 in terms of both maxi- mum lift and angle of attack at which maximum lift occurs: for both 6F - 300 fall slightly below the prediction based and 65 0 , the maximum values of CLi ► et on tunnel tests. Angle of attack for maximum lift agrees well for 6F - 30° but follows a different trend for 6 F - 65 0 : there is now z distinct variation A-6303

with blowing coefficient - values of aw for maximum lift are lower than pre- dicted at low CJ and higher at high Cj. Evidently, for 6F - 65% an increase In the blowing strength permits the wing to probe deeper into the stalled region while not generating a corresponding increase in lift.

The predicted nature of the stall has been considered in s..ne detail in reference 4. In summary, the tunnel tests suggested that there were three predominant factors; first, the entrainment of the secondary flow into the augmentor tlap provides a powerful means of boundary layer control at the mid- chord station of the wing; second, that the onset of stall occurs (quite pre- dictably) at the wing/fuselage junction but that the disturbance is confined to the wing root because of jet entrainment; and third, that the presence and growth of this; wing root disturbance causes changes in downwash at the tail to generate a post-stall pitching moment in the nose-down sense.

Life and pitching moment characteristics for the landing configuration are given in figure 16 for the large model fitted with pylon mounted nacelles.

Test results are shown at a wing blowin- -oefficient, Q - 0.85, and with of thrust coefficient, CT: 0, nacelle thrust vectored at 90 0 for three levels 0.33 and 0.82 (based on thrust from both engines). The very gentle nature of the stall is quite evident and so also is the nose-down moment which increases as the model is driven further into the post-stall region.

This behaviour may be compared with some specific flight experiments (ref. 5) with landing flap as shown ifs figure 17, where point q in the post- stall region represent a quasi-steady state (that is to say, the pilot was able to fly the aircraft well into the past-stall region in a progressive manner). The time history shown represen`s a typical stall for landing flap - 65 0 ) withthe hot jet nozzle vectored at 67". In this cakie, the post- O F stall region was maintained for 6 seconds whereas, in one similar case, it was held for 16 seconds indicatinp a relatively docile stall as predicted.

Corresponding measurements of elevator angle to trim show that ul.-elevator is required to hold the aircraft in . the post-stall region once again following the trend established in the wind tunnel.

fit of speed variation on the rig ttt of fig- The same stall is shown ure 17 where the large speed margin (about 8 m-2ters per second) and large of margin" between approach and the stall (about 20 degrees) is now evi- dent as discussed in reference 6.

The following comments made by ona of the project pilots confirm these impremsions and serve to illustrate some other aspects of the stall.

Flaps 65 0 , vectoring nozzles 60 0 , engine speed 95 percent maximum.

"The initial buffet was accompanied by a relatively mild roll acceleration to the left and an almost immediate shallow nose down pitch of not more than a few degrees. At pitch down, the indicated airspeed increases 6 to 8 knots very promptly, which made it difficult to continue further aft movements of the column, since appeared to have been reached and the speed/elevator rela- CLmax tionstip was essentially pest-stall. The initial left wing drop was Easily limited to a bank angle of less than 10' by coordinated use of right A-6303

the

roll control and rudder and a slight forward movement of column effected a prompt negative pitching response for recovery. When applying roll control to c..-tnter the post-buffet roll the roll control system rate limit perturbation was encountered which made it difficult for the pilot to stay in phase with the post buffet rolling motion."

Flaps 30 0 , vectoring nozzles 6 0 , engine speed 91 percent maximum.

"The predominant aspect of this stall was the fairly extreme nose high attitude reached prior to the onset of buffet. Longitudinal stability stick- fixed was more positive and a pull force was noted throughout. The maximum and following indicated pitch attitude reacted was something in excess of 32 0 , the onset of buffet the elevator could he further applied to the aft stop, following which the aircraft demonstrated a shallow nose down pitch. No

reversed response to pitch command was roted."

CrounO Effect Tests were conducted on the large scale model in ground effect at twc, heights, h/c - 1.3 and h/c - 2.1 for takeoff flap (50°) and landing flap (750).

For takeoff, thrust of the pod engines was directed backward in the conven- tional. manner, whereas for the landing configuration the pod thrust was deflected downward through 85". Data from there tests could he compared directly with corresponding drta for the model mounted at the centre of the tunnel (h/c - 3.5) where the model is essentially free of the ground plane.

At a ground height of h/c - 1.3 the wheels cf the research aircraft would be just touching the ground with oleos extended (fig. 18).

For takeoff flap, grc.unC effect on lift at h/c - 1.3 was negligible; drag was reduced slightly and some nose down trim change was evident. Once again, it is the landing flap configuration which is critical and deserves more attention.

Even the landing flap configuration showed little ground effect on lift tested (h/c - 1.3, and drag with zero pod thrust at the highest value of Cj = 0.85). It was the configuration w:ch vectored pod thrust - 75 0 , C ,1 1 6 F which introduced significant ground effect but even then, for h/c - 2.1, at mc,derate values of aw, the model exhibited classical ground effect trends - some increase in lift, a reduction in drag (at constant CL) and a change in downwash at the tail. (Not illustrated.) However, at h/c - 1.3 the trend was reversed with a reduction in lift and corresponding increase in drag as shown in figure 18. This result led to serious worry with respect to the flare and touchdown characteristics which were likely to be experienced on the research aircraft.

Clearly, this adverse effect was due to impingement of hot jet on the ground and the tendency for it to deflect forward and then spill over the wing.

However, there was some reason to expect that the tunnel results were pessi- mistic, because, first of all, the jet nozzle on the model was closer to the ground than for the research aircraft, secondly, there was a single jet at A-6303

each nacelle rather than a pair of smaller jets as on the research airplane, and thirdly, the presence of a tunnel boundary layer wil l always aggravate the situation by making it easier for the jet to be deflected forward against the wind stream ( in the absence of a moving belt to represent the ground plane).

It was on account of these doubts that, in t'ie initial stages of the flight trails, descent gradients were kept small and the flare manoeuvre was explored with some caution. As experience was gained it became evident that large ad verse ground effects were not present and that, In some eases there appeared to be a tendency for the aircraft to float. This lead to a series of tests in which the aircraft approached the runway with a very shallow descent gradient and was flown in a quasi-steady state only a few feet above the run- way. Results of this investigation are shown in figure 19 (taken from refs. 5 and 6) where it can be seen that 'lift, drag and pitching moment all follow the classical trend and that no serious adverse ground effects are present.

At touchdown, the lift is about 5 percent above the free air value.

t. NGITUDINAL STABILITY AND CONTROL Wti:d tunnel data showing lift, drag and pitching moment characteristics for the landing configurdt : on are s:iown in figure 20 (tail on, elevator angle zero, nacelle thrust vectored at 85°).

Certain broad conclusions could be drawn by simple inspection of the results.

- that longitudinal static stability would deteriorate as power Increased, but only to a moderate extent that changes in trim with power (at constant speed) would be - relatively small - that an increase in power would lead to a nose-up change in trim (e.g. to facilitate wave-off) - that elevator angle to trim would be small for steep approach at low speed.

on the left of figure 21, flight results are shown to confirm the expec- tations outlined above for landing flap. In particular, it is of interest to note that the aircraft is in trim with 6 P = + 1 0 during approach at 31 to 34 meters per second, which reserves the full range of up-elevator for flare and ijurhdown.

• Location of the engine vas chosen to minimize change in trim due to noz- zle vectoring for the wind tunnel model and for the research aircraft. The graph on the right hand side of figure 21 indicates that, in fact, this objective was achieved in flight.

A-6303

LATERAL STABILITY Consideration is given to static_ lateral stability in general and to dihedral effectiveness iu particular. A typical set of lateral data from the wind tunnel is shown in figure 22 for a landing configuration. Model tests were conducted with the addition of strakes fitted to the rear underside of the fuselage to make it more representative of the Buffalo aircraft. The model exhibited a reasonably good directional stability but the dihedral effectiveness was shown to be essentially neutral.

In contrast to zero dihedral of the model, the B-.tialo aircraft has 50 dihedral over the wing panels outboard of the engir ,-- nacelles, and zero degrees for the centre section. The influence of this additional element of dihedral was completely unknown, esp.cially, in comhi,;ation with large flap angles and wing blowing. For the simulation, dihedral effectiveness was varied between the wind tunnel value CZ g - 0 and a theoretically derived value CAS - .004/ degree to bracket tl.e likely handlit 1.,1 qualities.

Some results obtained from steady sideslip manoeuvres on the research be seen that the effective d1he- aircraft are shown in figure 23 where it c, , n lral corresponds to about Q R - -.003/deg, giving the more favourable char- acteristics anticipated for the research aircraft due to the outboard dihedral, but nevertheless, prior to flight, this aspect of stability did represent one of the major elements of uncertainty.

Making reference to '.he same figure it can be .;een that directional stability is close to prediction and that the overall characteristic is to give appro).'macely one Jegree of rudder per degree of sideslip with the con- dition remaining quite lineLr out to ± 15 0 of sideslip - this observation again relating to the landing configuration.

ROLL CONTROL POWER Toe approach and touchdown speed of the research aircraft is about 20 kt.

below that of a etandsed Buffalo aircraft: nevertheless the size and control power of both the horizontal and vertical tail were considered to be adequate (the cross-ducting begin a major consideration in this regard). However, it was recognized :hat a significant increase in roll control power would be required to give satisfactory handling qualities at low speed and, accordingly, special attention was given to this requirement during the wind tunnel test program. A roll control system was devised which consisted of three elements, a blown aileron, a spoiler located just ahead of the aileron and a small flap which reduced the exit area of the augmentor f'.ap in a region outboard of the engine nacelle. This latter control t)ecame known as the augmentor choke.

A typi.:,. set of data taken from the wind tunnel (fig. 24) shows the characteristics and relative effectiveness of each element. Individually it 10 A-6303

can be seen that each form of control hobs effectiveness throughout the com- plete rang y of C L and each is reasonably linear over the range of control tested. In addition, tests were carried out with the various controls in com- bination to assess interference effects. These roll.ng moment characteristics were used by Boeing in the design of a powered contro l system (using all three

elements) to generate the set of flight data shown it figure 25, giving a high

degree of sensitivity about neutral control deflection and high rates of accel- eration for maximum control. Maximum control deflections chosen f,-r the research aircraft were ailerons ± 15°, spoiiers - 50° and chokes 55 percent.

Correlation of roll control power between tunnel and flight is illus- trated in the same figure Uy a single point prediction based on data from reference 5 a maximum wheel angle. It is evident that the flig'at result came close to expectation in this respect.

Although not relating to correlation, an important design feature of the research aircraft is that blowing air is cross-fed to the wing so as to mini- mize moment imbalance due to vectored thrust in the event of engine failure.

In the approach configuration, roll imbalance is very small so that maximum control power is available for manoeuvre and for SAS actuation. For takeoff, engine out yawing moments are compensated so as to permit takeoff on one engine from a standing start.

FLIGHT SIMULATION TESTS Slight simulation was carried out during the design phase on the large six-degree of freedom moving base unit at the Ames Research Center kno , .n as the "Flight Simulator for Advanced Aircraft" (ref.7). Wind tunnel test data from the large scale model provided the main source of aerodynamic derivatives and characteristics for the simulation whereas rotary derivatives were esti- mated theoretically. Variations in ground effect and dihedral effectiveness were introduced to bracket the degree of uncertainty in these two parameters.

:he cockpit of the simulator was modified to make it closely representative of the Buffalo aircraft with regard to the instrument panel and location of controls. In particulat, a nozzle control lever fo-- thrust vectoring was L_ Lited right along side of the overhead throttles of the standard Buffalo.

The handling and control of the research aircraft was examined for all operating modes such as, approach, flare and touchdown, transition and glide path intercept, single engine failure, etc., and, over a wide range of atmo- spheric conditions such as wind shear, cross-wind and turbulence.

In retrospect it can be said that the simulator provided a reasonably close approximation to the control and handling qualities of the research aircraft. In fact, following the initial flight, one of the first comments by T. Edmonds (Chief Boeing pilot) was to the effect that "the airplane flew well at all speeds with handling qualities similar to thoso experienced on the • simulator". Subsequent experience has confirmed this view with the exception of the landing manoeuvre. It is well know that flare and touchdown always prove difficult to simulate for a variety of reasons; the situation was A-6303

aggravated by the adverse ground effects dictated by results of the wind tunnel tests. Even with zero ground effect set up on the simulator, flare and touchdown was not easy. As described earlier, flight experience revealed a favourable ground effect which resulted in a more accurate and gentle touch- down on the aircraft than was experienced on the simulator.

Earlier indications from analysis of flight results suggest that the estimation of rotary derivatives was reasonably good - in all probability, this contributed significantly to the success of the simulation.

CONCLUSIONS Data collection and anaLysis•of wind tunne: and flight experiments are more difficult for a powered lift type of aircraft because of the introduction of blowing coefficient as an additional major parameter Yid because of the pow,rful influence of thrust vectoring on performance. N.,twithstanding these difficulties, many aspects of aerodynamic performance and stability have been investigated and a reasonable level of correlation has been demonstrated between wind tunnel anti flight test data.

In areas were some uncertainty did exist because of gaps in the design data base of differences in geometry, the research aircraft has come out mostly on the good side of the ledger. This must be attributed partly to good fortune.

ACKNOWLEDGEMENTS The authors wish to express their appreciation for assistance in analysis of flight test and wind tunnel data from H. C. Quigley of NASA, Ames Research Center and from K. A. J. Lockwood and J. E. Farbridge of de Navilland (Canada).

The wind tunnel and flight test work reported here form part of an inter- national research program funded by the National Aeronautics and Space Admin- istration, the Canadian Defence Research Board and the Department of Industry, Trade and Commerce of Canada (DITC).

12 A-6303 I

REFERENCES 1. Cook, A. M.; and Aiken, T. M.: Low-Speed Aerodynamics Characteristics of a Large-Scale STOL Transport Model with Augmented Jet Flaps. NASA TM X-62,011.

2. Ashleman, R. H.; and Shavdahl, H.: The Development of an Augmentor-Wing .let STOL Research Airplane, Volume 1 - Summary. NA Q 4 CR-114503, August 1972.

Williams. J.; Butler, S. F. J.; and Wood, M. N.: The Aerodynamics of Jet 3.

Flaps. Royal Aircrafc Establishment, Report no. Aero 2646.

4. Whittley, D. C.: The Augmentor-Wing Research Program, Past, Present and Future, AIAA Paper #67-741.

5. Quigley, Hervey C.; and Innis, Robert C.: A Flight Investigation of the STOL Characteristics of an Augmented Jet Flap STOL Research Aircraft.

NASA TM X-62-A4.

6. Cook, W. L.; Hickey, D. H.; and Quigley, H. C.: Aerodynamics of Jet Flap and Rotating Cylinder Flap STOL Concepts. ACARD Fluid Dynamics Panel.

Paper no. ;.0, June 24, 1974.

7. Quigley, H. f.; and Holzhauser, C. A.: Requirement for Simulation in V/STOL Research tircraft Programs. ACARD Fluid Dynam cs Panel. Paper no. 25. June 2'&, 1974.

A-6303

'TABLE 1. - Moi AND RES1 ARCH A I RCRAF°1' GEOMETRY Wind Tunnel Research Model Aircraft Wing in Gross area, sq.

80.36 Span, m (B) 13.46 24.0 7,2 Aspect ratio 8.45 in Root chord, 3.77 Tip chord, in 0.81 ;1.36 i'hicknebs/chord ratio .16 .175 - .15 Dihedra_ angle 0 5.0 Flap Chord, m (aft of hinge line) .56 .91 Chord ratio .33 .24 Span, in 3.72 7.01 Semi-span location k(b/ 2 ) .12 - .67 .12 - .72 Aileron (^] Root chord, m (aft of .4 hinge line) Tip chord, m .24 .61 Span, m 1.72 3.50 1.0 Semi-span location k(b 2 ) .68 - 1.0 .72 - Spoiler Chord, to .17 - .10 .36 Span, m 1.52 3.44 .72 - .99 Semi-span location k(b /2) .73 - .96 Horizontal Tail m 5.92 21.65 Area, sq.

Span, m 5.27 9.75 4.4 Aspect ratio 4.68 14.11 Tail arm, m 7.25 2.42 5.0 Tail height, m(rclat?ve to wing) Tail volume, V H 1.16 1.0 Vertical Tail 14.1 Area, sq. m 3.9 4.14 Span, m 2.30 1.37 1.2 Aspect ratio 6.25 13.23 _ Tail arm, m .085 .097 Tail volume, V v A-6303

TABLE 1.- MODEL AND RESEARCH AIRCRAFT GEOMETRY - Concluded Wind Tunnel Research Model Aircraft Angular Settings Wing incidence, deg 00 +2.50 (relative to fuselage) Wing dihedral, deg 00 +5.00 (outer panels only) Horizontal tail setting, deg -40 (relative to fuselage) A-6303

W A -i- v Lai

` o

a R1 f+.

O N b u L y a v

. z

C)

m w C.3 4

I

W Z

W

' XP G I'll ^ PAGE BLANK NOT

J

1.

:J G C r :s

J

N

ro

:i

n

v

P4 0.

ro 4.

u

"O Cl u F Q^ b0 -L J :.0 W

N

w

u

w

,r d

u

v

N y PG

a

ro t^.

b o^ a.+ G Qr b0 G d M

w

's a

O

a IE

SIL

CLIo

W 0 a a

a

v

a^ G a '.0 a G I a 1.

•.ti t+.

O

tot;!

s

_C_

c=.

ac I . - cc ci a G V

i

I ZIP

Z O

cc

to

L W

O

TE

I C9 O

r H W O O

J

W W

m

Q ro O

^J a

F SC CD m

O

a II CD z CIO b x y

^ Z

O

C3 Lei Gn QqKT o O ^ O S CD w O = o qdl ~ .^ d ^ V d W `^ G cr- O C-) w J W ► ,, = W C=) O °' a W C=) V O C11 4 q W N V) a w to .,, LL.

CD O O

O O O O 0 cm CD CD CD

CD O O O O O N N O OD cc co ca qqr N N ) Cl) cn !h

N N N N

H

N

CY m S J H m cz x z_

O

J m

I

J W O Cl

a

V) W N ^_ H b0 N Ll-A G = O

J

co C •N N u ) CY N O u ^ O CC H U = END C G o

V1

Or. 1ll .Q W cci W G O Z W n Z s~ I I 0 N O = ^n ca Z v W W Z ,J L, LJ F

s

V J z W L n Q C G O .o b4 H h

Z r

W

G[ Cp a = to F- C) = O .Now r, ► 1'1 cn v W N

PV O

0 GC

^O H U

CL LO ac

z

v 0

O

U ~ U C ^ o0 3 1 a O CL F W .a = LQ W G I"- CD V v a 00 J W W

O (}I

O IE

W a

x

Co r v W Z Y^ V V I ti / I G] C" I ex s P` ^ GG DC o W G ^ Z G tV I V C

N I I

M I •- t of e 11 w \ \ _ u^ v \ \ a u•i Ln u w u

ca

Z ^ \ ^

O

rcri I ^, U v, c 1+ N v v I Z ca ac

x

r^ I O W I ^ V I ^ I M W th J I U H M a ^ N h I

h

f 1!

w 1 p

O

H

W

I I

at

cc CL w o 3E II

W

II W a C.3 c

h J

c n J W 4 = V ^ II d co ^- Z S H W Z

m

a

r— u G O tir _ co an V)

y

m

W a

tt W

Ln O

N

O W

Ln O u ro Z rra.

G

c4 L3 CC

G W

O

as 1=- __ y N Z W W N Z C . Lai u ^ J yCD G 22 = J O O

N

Z

O N

Q b^G

t

L&i O

O d J

d

Z CW) Q dw P14 JJ

J

LA-

W O

cr. Z Ca H

y = W

CZ)

N

Lai G cr.

I

CL ; O

rn

v

J

dc W

O

N J Z

IW w (a •,a

O_ _ _

I O O

M

C14

Z

O ^

H

d '

W ~

Q0 CD O CD C-3 ca co CD O

O

3 (M

C" C.

C*4 Ln CN LL.

em,) ci U') C= cn C6 If II O C4 CD O LAJ O CL.

ar.

ts LAJ LAJ CD CD C3 IM 2c LAi ocr.

FE

C=) I— CD W2 qr C14 I.— LAJ 2c

d

m

CD

J H

G

Q M Ln 1l7 U'! V) N O Z W ^"^ to GO f^ O LAJ .D g o cl CO - -- -- W W UI J C O I N V N D n 00 ^I p O N C w N W D Q = O O 4 d O d o G ^+ _._ . ^? _ c ^ O O s C d II w

3 ^

v b w i G v v - ^ O u In V I— M N ^ O to w W w Z 01 O J U

V

u w N tA

a

= Z O

w cm cm Z Z W W G O O N N LQ >J O

a II ^

CD

O

m m

V

v

W

2E ZE ~ ^°

U > }'

N

- V) NJ -

O

W Z

cc W J v CL CL O H O N O ac O w U-3 co cn N ^f

O

H W Z J V

rn O

w

_

o

0 V

N M N

M

-

O-

N

cl

1n ~

W W C Z O O Z H

^ V

► ^ N O W 0 Ln W ^ I M ^O O a" S 7 W W J

Z

N D ^. k.

- Z

w

H

N O

O

t7 O a

^ Ln 1— c+)

N O

W GO N

Z

1+ J Q

V

w O G O N

O

H N

O

a m e

^_ -

J G

o

W U

Z

f^ I

Z

' - —t - O

O

N

co

a 3

-^ o y

O

Z V

C O 3 W N I.- Lk- t+.

O ^ V

2E V

W W D

O

.a Z at

C6 v

c c a

O U cn W 51. 4

O

Q C39 W

N

p O

C

O

O

O

Ili lw M N W p

W

Z

J C.^

^

m

o' 0 qw

M O ca

V

O

II G.

ti O O .^ G o .r+

V

O a^ O

N

u O v .b

u

G a -1 O u O C) W- O N w

H

w W 6) Z O

J

U

V

L

W y.{ ra cc O W G G O y O tw O a O d O

c

C.3

O

u i0 - I

n

3_ O

O

^ O u w

w O

O V— qr M N

O

Z

J

V

W Z ^ W V

O

LAt O O o d d O l! N

O p

cn 1 1 O O I am CO N w w F- ch ^O O r y H op W J O W O P.

J Z cm at oa Z W N H Q J Q O N w Q N W

0 O

J

I a. Z

G Cl O V LO) .,a M O Cl.

O U

d

G1 Gs.

O I N O I

T C14

—

I

-- co

cl • o O U

cc

ca

= W U 92. r-4 tD O t C v - G u H W Z C C^ • O W O w 4+ fs.

LAJ N J Z w N H O ► .7 V O E I 3 G:3 O Ll1 O U W Ch N N

O

x • CD Co n f^ W- CD Q Cn N CD x LC!

X

a

w O

, J

G V O O y r O O

W

O^

ro

ch Z Z

a O J Z W O CD -- - U

J

O J ca C^ Z O co I

LLJ O

Z Q

Ln CD

-- -- a N 1d -- -

CD

v o

PQ LLJ n 0 O

Li Z

na C3 CL w W

O

I Z C3 N O I ti LIJ Ln Ca !Or! N N 1^ CD LC2 it m N O K Q " J

w

w r^ w O O

O V

^rl u c^ O 3 % W L r.

cJ IL') O 00 1e^ r O 0 Rp qp M N ;s he J O O O'1 U.) um) V Lr) co CO C: C m v W O J x cm Lai J= OG CD Q ^.

t Z Uj

4 J ^ ;j

N rl N Is.

.J

L z

O ^n

G N

ro W

L

C7

W

.,4 W Z

^ U

_ 3

y O V

i

I

L

O d H O ^.d O

w

co C)

o I N

b H _.)

C)

M

N

O O 7^D Q M V

^ N

w

]C

N

11') t

N

c p 0C = O ^^ W O ^ N cl F J Q J W ,^ ; ^ CD ---- LL7 O r.^ C" O a a ro CIO s ^_ N pa CM

Z .v

cx G W u -i Ln W

a

u N O W ~ O co co N 2 = CG Ul) Ul) ^ O C~ ^ 1~ L>D 1+•

Z W

= W Y ^ u W LID W y w J d 1t N u ^ >r iE

z

a

M

CD O

^C U 0 ~ O O

OC Ft

v LO r— v1 ca co c W O 66"1 J p / w Lai = 1 u'f = ??

C3 ^t n W " C - 0 Q Lai C:) O W

Q O

v

J - N

N ~ V 3 I s.

CL N H- \ \ Q

J O J

1i 2 Q C14 O

O p G

^O O f l- cc 1A N .-- p H J V

ri

U

u

1L7 of a O v1 ^ O ^ ^

NC

a

a

CP t0 1!f +^ !r! IM li N

o I

v

^O

,^ G

G o ~ W.)

N V L ^ O O H Cn aD IA II m n G = II ^ II LL.

H- ^ 1^ ^A F F — Q o Z O W C.7

W N Z w

O J W O > L LIJ 0 C.

(_M L^ V J l u CD

' r

w 4K cm c0 Z W Z W W ^; a 4 J ,J y

Z

LAJ N NJ i O Q i N p N ; 4. Z m Z CM a v ► 4 O w ^I h- w ^

J

c^ ci L•. La.

J Li' W G)

U

at

0 ? N

W

J

G

O W Z O N O W W ^ ^ O V o O ^ d OD O C:, G CO N iC to L 1+.

u w s+ Q w o +a o CD G ^ oc%) co C.3 ca O aC

CC! d

O C) p U ^

a

U

o =

J Z

W

tO .d O o r v O L w O O u OD u v ^ — w O Lr) 4.4

I q

co N'7 Ln Ln q W O to J N cn ^ f- q i

r l

n o v

n _

`r

ca q ^ CA OJ W q d J V V ^La N IL N Q O

J

LL, Z

o O co

i0

N O

1— p

S

S

Ca O w

--- W =

=

V t

Co O

Z Z

al:

s

W

a O

U

CD

O

J ^ N

s A !A - soft -- i -- r `t

al r 1 (r

U

rl

N

a CN Go F.

O

i

II GI G C M..^ co a E- c^ ^v

C

a cm • u a -a W co ^ so to IRr cr W— O 1 a^ N

J

n -1- CO) U) ca Ih-

U

00 CO M Z N - O 1+ O

U O

G+

^ 0 0 0

N u O ^o

C O O Q D

m~

J

N

CJ

M

C) G Z

- O .O

N O

O

u

LO M C' G

W

O - O

D

W

r"i I V

a

Q

M

cJ N Q M r C+r L` C n OD 1'+ CD

1!! m N O

J

I

I Cl N U C—c

V Wfl-

N G N a H

E- -

0 ca

LA- Z x ^ eA Coc o GO Z

z

w Y

4 Go D

O O

W W W O W W OD ?

i - J —^ a.

W w = cn P%4 P"4 CD w ca Z

a i

z

Li- W V O 0!

x Co to N d N qw tC ra I cm I I I iy Q r, a^ ,G tq u O

N

b

I

I J-

c^

O

0 0 G ^r P- O

Z

O C" `^ii CO H

O N

4-

W ^ r, ch LA- x C" 0 G `O x O ~ b GO a W u

ZJ

O ' J cc W ^^ W ca C

Z C a

aW I

Z. 'J

CD

W C N

N J ^ Lt C!

1+ Oq N O c-4 CO cc r' ^p co Cm W

v

n^ I cq

to

N

co qw O II N W F-- V ca

V

cz CD

N

LA-

m

W C O O N v.

LC) O u OD U') a F

V1

C N r V dm V1 LLA cr.

9w = W W d Q G Ca W W J G W a J J 1- H D P-4 N N Z ro P%4 N W G C co

Cl

O 's

Z Z

m

J 3

a

N

t

CD

Q

Z

v1

«4 r

O O C" 'J3 C^ W LC) Lan lD J r— 0 u cm cc f Z Nd II de

D

J O WE cm Z v

GL

U

Z L) a.^ W N Q cc Z CL Z qw Q W

V

N

V I-- N

LL N

Z LL v W LL O W O 00 Cl

N

w

V

C'3

Z

cz J Z J

O CD

at Go CD co CIO Q a-" %-4 o

p O O O

V^l

C C O O

1 1

r W Cm J - - ^ a cm 49 o t`

N

to J 1.J J to CD a H co CD ^c I W O 1+.

O O 19 O 117 O 119 O L17 O

u h .^ w ro 117 Na t~ ^ W u W tv 1 O d W

G

W v

CL

J N

_W —

O pC co ro v

z ro

v

J

d a

D

O W d W J J C.6

cm

W

U ..)

i

D 1 ^

m

•.y O ^ FS ^ W co I Q Ne u I c7 Ln Q 117 ^ W .J O O Ln O lf7 O 1171 O 1!9 O ro N ^ ^ 1 r' .; N

u

ro ri cv

W

O

J C/

$4

O

z d .,4

O

L+.

J W W i A O W L^1 O N 119 O 1171 O O O 117 O 1 C-O

o +' v H o' O d CO G ^ G d ^ G C G

In

Q

O

CL ;3

V O O 4.

N y ^L O W_ i.

L' II

J

U

M H W- O p C

r- m

J

♦ J o CD Z CD W o ^ O o i O I m O Q a

Ca

O W W n O

a

O W

O I

p

a.

i l t^/^ -- ^ O O ^ , h V o

U

U

N

-- - — ci ^ 1f7 C O 1 ► fD qr M N .^ O

J

rx

v

CD o fJ c.

1/9 Gl C O W ^ OC 1 C YJ c+.

L ^ IA O f^ ^O IA ^ M O p N 9- i J V

^

Y. A cc

a r

W Z f- O u ca 0 J W r/ 09 = CL 4- Z

W

W CD ]C CD -

M T

Q ^y.^ O CD

T W u

Y V- W J a.i oc

Z

Q

w W D u w A J R1 W

N

u

N

w cz ^O u a _

^ W

U M ) y.

v J N c" o W C3 a Z I M Co O ac Z O C!^ GO H d u n N 4 J ^ W O W 's

^ Z Z

S 0

a

V

J O Z

C^ 4 = a

p <n o u G

Q

O

V

U CL d J CD 0

L..^

a r A

N ch

qlt UM) O Q O

C

A A O

N

v = N ti O C) O F- W ao N1 G+.

W ^

J Q

W

V Q •^ J J Q