Applied aerodynamics: Challenges and expectations

: cii— NASA Technical Memorandum 103963

/1

Applied Aerodynamics:

Challenges and Expectations

Victor L. Peterson and Charles A. Smith

N93-2 5091 (NASA-tM-103963) APPLIED AERODYNAMICS : CHALLENGES AND EXPECTATIONS (NASA) 20 p Unci as G3/02 0158517 February 1993

NASA

National Aeronautics and Space Administration

NASA Technical Memorandum 103963 & Applied Aerodynamics:

Challenges and Expectations

Victor L. Peterson and Charles A. Smith, Ames Research Center, Moffett Field, California February 1993

NASA

National Aeronautics and Space Administration Ames Research Center Moffett Field, California 94035-1000

APPLIED AERODYNAMICS: CHALLENGES AND EXPECTATIONS Victor L. Peterson' and Charles A. Smith" NASA Ames Research Center Moffett Field, California Summary ingly stringent environmental regulations governing air- craft noise and engine emissions, maintaining acceptable Aerospace is the leading positive contributor to this safety as the number of aircraft in the airspace system country's balance of trade, derived largely from the sale increases, and producing aircraft having competitive costs of U.S. commercial aircraft around the world. This power- of ownership and operation. Military superiority requires fully favorable economic situation is being threatened in aircraft with unmatched performance for both defensive two ways. (1) The U.S. portion of the commercial trans- and offensive operations.

port market is decreasing, even though the worldwide market is projected to increase substantially. (2) Expendi- Meeting these requirements in both the economic and tures are decreasing for military aircraft, which often military areas depends on a strong aeronautics program serve as proving grounds for advanced aircraft technol- that will produce a continuing stream of technical ogy. To retain a major share of the world market for advances in applied aerodynamics. These advances can commercial aircraft and continue to provide military air- only be derived from innovations that emerge from basic craft with unsurpassed performance, the U.S. aerospace fluid dynamics and applied aerodynamics research efforts.

industry faces many technological challenges.

The U.S. government has been supporting aeronautical research and technology development for over 75 years The field of applied aerodynamics is necessarily a and without this government investment, continued lead- major contributor to efforts aimed at meeting these tech- ership in aeronautics cannot be maintained.

nological challenges. A number of emerging research results that will provide new opportunities for applied The two principal reasons for conducting a strong aerodynamicists are discussed in this paper. Some of these program in aeronautical research and development will be have great potential for maintaining the high value of discussed in this paper, and some related challenges for contributions from applied aerodynamics in the relatively the field of applied aerodynamics will be identified. This near future. Over time, however, the value of these contri- will be followed by highlights of some emerging research butions will diminish greatly unless substantial invest- that will contribute to advancing the state of the art of ments continue to be made in basic and applied research applied aerodynamics. In summary, the overall intent of efforts. The focus: to increase understanding of fluid this paper is to, first, show why it is important to the dynamic phenomena, identify new aerodynamic concepts, country to continue to invest in aeronautics, then to iden- and provide validated advanced technology for future tify challenges for the discipline of applied aerodynamics, aircraft. and finally to present some emerging research results that will help meet these challenges.

Introduction Economic CompetItiveness and Related Applied The United States should have a strong aeronautics Aerodynamics Challenges program for two principal reasons: (1) aircraft sales con- tribute importantly to international economic competi- The aerospace industry is a significant contributor to tiveness, and (2) superior aircraft are vital to a strong this country's economy. In 1991 it produced the largest national defense posture. Economic competitiveness in trade surplus of any sector of our economy. Values world markets for aircraft sales requires meeting increas- obtained from Ref. 1 for the industry's imports, and both civil and military exports, over the past 19 years are pre- sented in Fig. 1. These data show that trade surpluses have ' Deputy Director, Ames Research Center. AIAA Fellow.

occurred for a number of years (with a high of $30B being "Executive Assistant to the Center Director, Ames Research reached in 1991), and that the contributions from civil Center. AIAA Associate Fellow.

to age and to more stringent community noise restrictions.

CMI .xports Up to 300 aircraft each year are expected to be retired 40 Military .xports • through the year 2000, and 150 each year from 2001 to o Imports 2010. In contrast, only 91 aircraft were retired from commercial airline service in 1991. The net result of these EXPORTS forecasts is that deliveries of new aircraft will average A.rospac.

about 600 each year through the year 2010.

balanc.

of trads f 24 (.xports U A concern, however, is the continuing erosion of the minus worldwide market share enjoyed by the U.S. manufactur- ers of commercial aircraft. There are, of course, many 1e reasons for this decline in market share. One is that other a U 12 countries continue to make substantial investments in U aeronautical research and development and now offer C o8 products that are technically competitive. Another is the sometimes-lower cost of ownership of foreign aircraft made possible by lower prices due either to lower produc- o tion costs or to more favorable product financing. Addi- 4 73 .75 .77 tional loss in market share would have a severe impact on 81 83 the economy, especially if market projections for future 8 - commercial aircraft are realized. Capturing an important

•UL

segment of the expected future market will most likely IMPORTS 12 - 91 require development of a new supersonic transport, pro- 16 - vided current studies confirm environmental acceptability.

This implies an even greater investment in new technol- F] ogy than would be required to meet the challenges associ- Y.ar ated with a wholly subsonic transport market.

Fig. 1 Aerospace exports, imports, and trade balance.

This country originally gained preeminence in aero- nautics and the dominant share of the commercial aircraft exports have been growing steadily in recent years. It may market for two reasons. One was early recognition of the be difficult to maintain this growth in exports relative to importance of aeronautics, and the second was the gov- imports in the future because of the expected decline in ernment commitment to sustain investment in research expenditures for military aircraft. More than 25% of the and development. This commitment has existed over the nation's investment in research and development has been past 75 years, beginning with the creation of the National made in aerospace, and the military arena has served as Advisory Committee for Aeronautics (NACA) in 1915.

the proving ground for much of the technology resulting Halting this erosion of market share will require continued from these expenditures. Therefore, without a strong mili- substantial investments to find cost effective technical tary thrust, more of the burden for developing and validat- improvements that meet increasingly stringent demands ing new technology must be assumed by the civil sector if on performance and environmental compatibility.

aerospace is to be maintained as a leading contributor to the nation's economy.

Investment made by government through the NACA (and now NASA), and by industry is augmented by that Despite the decline in demand for defense products, made by the Department of Defense (DOD). Much of the economics of the overall aerospace industry appear strong aeronautical technology developed with DOD funds over because of the very large increase in demand for new civil the years also has been useful to the civil sector. There- aircraft. There are two reasons for the increasing demand fore, any reductions of DOD investments in aeronautical for civil aircraft. The first is the expected increase in technology must be compensated through greater invest- annual revenue passenger miles (RPM). Data presented in ments by industry and by government through NASA.

Fig. 2 (taken from Ref. 2) show expected increases in Finally, if a product is not less expensive than competing RPMs to be 70% between 1990 and 2000; more than products, it must be technically superior and clearly double by 2005; and almost triple by 2010. The second satisfy user requirements to compete in the world reason for the increasing demand for civil aircraft is the marketplace.

expected increase in retirement of older aircraft due both

C 0.

0_.

1970 1975 1980 1990 1995 2000 2005 2010 Fig. 2 World revenue passenger miles, past and future.

Advances in the field of applied aerodynamics are Engine emissions can be made more acceptable by crucial to both maintenance and improvement of eco- improving ways of mixing fuel and air in engine combus- nomic competitiveness. Some elements contributing to the tion chambers. Also, adverse consequences of emissions economics of an aircraft that can be affected by applied on the Earth's ozone layer can be minimized by designing aerodynamics are pricing, direct operating costs, envi- aircraft that will cruise efficiently at the desired Mach ronmental compliance, and safety. number, at altitudes not much higher than 50,000 feet.

Reduction of engine noise involves design of low-noise Competitive pricing of aircraft depends, in part, on combustion chambers, light weight exhaust-noise sup- manufacturing costs, which are influenced by the degree pression devices, and inlet and compressor designs to of design complexity. This challenges the applied aerody- minimize compressor noise. Noise can also be minimized namicist to find simpler designs to manufacture. For at communities surrounding airports through development example, simpler high-lift devices to perform the func- of aircraft with very high lift at take-off and landing tions of the sometimes quite mechanically complicated speeds. This would both keep the high-noise footprint on wing leading- and trailing-edge devices. Also, minimizing the community as small as possible and permit operations the number of aerodynamic surfaces and keeping them at lowest possible levels of thrust (low exhaust velocities).

simple, both geometrically and mechanically, will help to The challenge to reduce airframe noise throughout the reduce manufacturing costs. flight envelope calls for minimizing turbulent flow and the regions of separated flow over the aircraft.

Direct operating costs are strongly influenced by the amounts of fuel and time required to perform a given Safety can be influenced by applied aerodynamicists mission with a given payload. Fuel required can be min- in several ways. As aircraft become larger and faster, imized by maximizing the Breguet range factor. This calls landing and take-off speeds have a tendency to increase.

for maximizing the product of cruise Mach number and This must be overcome by designs having high lift capa- lift-to-drag ratio, divided by the engine efficiency bility at low speeds. More and more people traveling by expressed in pounds of fuel per pound of thrust per hour. air, and the very high cost of constructing new airports, Challenges for applied aerodynamicists are to maximize are both causing the number of airport operations to reach aircraft drag-rise Mach number and lift-to-drag ratio safety limits. Some of this congestion can be relieved by (including the installed and operating engines), and to designing new aircraft that can carry substantially greater work with the propulsion engineers to find more efficient numbers of passengers. Additional relief can be obtained engine inlet, nozzle, compressor, turbine, and combustion by finding ways to reduce unfavorable effects of trailing chamber designs. wake vortices from a leading aircraft on another aircraft following closely, If this can be done, spacing between Environmental compliance is largely influenced by aircraft can be reduced safely so that the number of oper- engine emissions, and both engine and airframe noise.

ations handled by an airport can be increased. Still another

way to relieve congestion is to develop short-haul aircraft, economy and aerodynamically efficient airframes. In having either very short or vertical take-off and landing addition, weapons and stores must be more highly inte- capabilities that would not have to compete with long- grated with airframes than has been the practice in the haul conventional aircraft for main runway take-offs and past. A challenge is to increase the level of integration landings. while still retaining capability to rapidly jettison stores (drop tanks) and deploy weapons. Increasing speed into All of the challenges identified above must continue the hypersonic Mach number range imposes requirements to be addressed by applied aerodynamicists if manufac- for applied aerodynamicists to consider real-gas and, turers are to remain competitive. Major advances are still sometimes, low-density effects. Aerodynamic heating also possible in each individual area. However, by more tightly alters the behavior of boundary layers and, consequently, integrating a number of disciplines in the design process, the friction drag component of total aircraft drag.

even greater advances will be possible. This will require simultaneous consideration of disciplines such as aerody- Low observability requires special attention to both namics, structures, propulsion, and active controls. aircraft shaping and noise suppression. It is well known Applied aerodynamicists probably should take the lead in that this imposes requirements to seek radical new ways forcing this increased integration. In each individual dis- to shape an aircraft while still maintaining reasonable cipline, computational approaches will provide the com- aerodynamic and propulsive efficiencies. Over the years, mon interfaces necessary to make all of this possible. applied aerodynamicists have developed many empirical factors for designing conventional aerodynamic shapes.

These can no longer be used. In fact, aerodynamics can no longer be considered independent of electromagnetics Military Superiority and Related Applied Aerodynamics Challenges well considerations. Rather, both of these disciplines, as as materials considerations, must be treated simultane- National defense has become increasingly dependent ously in the design process.

on capabilities to conduct air warfare and to move large quantities of people and materiel long distances in short Extremely high-altitude subsonic flight poses another periods of time. Maintaining superior capabilities in both challenge: designing an aircraft with very low wing load- of these areas requires continuing development of all ing that is also manageable in gusty weather at low alti- types of aircraft including air-to-air and air-to-ground tudes, particularly during take-off and landing.

fighters, strategic bombers, conventional transports, short take-off and landing transports and fighters, manned and unmanned reconnaissance aircraft, and rescue aircraft. Emerging Research Results and Their Implications for Applied Aerodynamics Most elements discussed earlier that contribute to the Emerging research results within this decade will economics of commercial aircraft also apply to military aircraft. Additional elements, either primarily associated contribute greatly to meeting many of the challenges for with military aircraft or requiring additional emphasis, maintaining economic competitiveness and military supe- are: maneuverability, agility, speed, low observability, riority. Since many of these results derive from computa- and service altitude ceiling. tional approaches, a brief description of equations governing aerodynamic behavior is given.

Maneuverability and agility call for removing restric- tions on the attitudes in which an aircraft can routinely operate. One challenge is to make fighter aircraft stable Equations for Aerodynamic Behavior and controllable at all angles of attack and for reasonably large ranges of sideslip. Another is to provide the capabil- The Navier-Stokes equations describe behavior of ity to develop side forces without introducing yaw. Still fluids in the continuum flow regime. These equations another is to maintain the capability to develop high pitch were derived over 100 years ago and now they can be and roll rates in any attitude. These challenges require solved, without approximation, for flows about complex new ways to control an aircraft, either solely through three-dimensional aerodynamic shapes. However, results aerodynamic means or through combinations of aerody- of estimates presented in Ref. 3 show that computer namic and propulsive means. requirements are now excessive and probably will be for many years to come. Fortunately, approximate forms of Sustained flight of an aircraft at very high speeds the equations have been developed that can yield excellent obviously requires high-thrust engines with good fuel

predictions of aerodynamic behavior for many engineer- Boundary Layer Transition ing applications.

It is difficult to accurately determine the true perfor- Five different levels, ranging from the most complex mance of an aircraft before it is actually built and flown.

full equations to the simplest approximate form, are: One contributing factor is uncertainty in the location of (1) full Navier-Stokes or direct simulation (DS), (2) large boundary layer transition from laminar to turbulent flow eddy simulation (LES), (3) Reynolds-averaged on the various aerodynamic surfaces. Two factors that Navier-Stokes (RANS), (4) nonlinearinviscid, and contribute to variations of transition location between wind tunnel and flight measurements are Reynolds num- (5) linear inviscid. Solutions to most problems of current ber and freestream turbulence. Tests of subscale models in interest in aerodynamic analysis require consideration of wind tunnels generally cannot duplicate true flight viscous effects. So considerable future attention will Reynolds numbers, and wind tunnel freestream turbulence likely be placed on either the full Navier-Stokes or the is usually higher than that in free air.

LES equations for benchmark research studies, and on either the LES or the RANS equations for practical It is possible to calculate the location of transition applications. Brief descriptions of the LES and RANS from first principles by solving the full Navier-Stokes approximations follow.

equations. Previously, this had been done only for very low Reynolds-number flows, particular geometries, and In the LES approximation, large-scale turbulent non-arbitrary upstream and downstream flow boundary eddies that can be resolved by the computational grid are conditions. All of these were dictated by limitations in calculated directly, and only not-resolvable small-scale motions are modeled. The premise of LES is that large- computer speed and memory.

scale motions carry most of the energy contained in the flow, and essential characteristics of turbulence dynamics Recently, the numerical method described in Ref. 5 can be captured from these large scales alone.. Only the was developed to treat general geometries and arbitrary small scales of turbulence, which are presumably more flow boundary conditions with computers now available.

isotropic and universal in character, are modeled. Initial results for the skin friction along the length of a flat plate are shown in Fig. 3. They are compared with mea- The RANS form of the governing equations neglects surements made in two different investigations (reported no terms in the full equations but all scales of turbulence in Refs. 6 and 7), as well as with skin friction distributions momentum, energy, and heat transport are modeled. Many from laminar-flow theory and from a correlation (pre- different problems are now being solved with these equa- sented in Ref. 8) of measurements from numerous tions by using currently available supercomputers. Com- turbulent flows. Flow Mach and Reynolds numbers at the puting times range from several minutes to tens of hours, leading edge of the plate were 0.1 and 50,000/in., respec- depending upon complexities of geometries and flow tively. Freestream turbulence intensity level at the plate physics. Unfortunately, no "universal" turbulence model leading edge is about 2.75% for both computation and seems to reproduce "true" flow physics in all situations. experiments. The computed results lie between the two Some models provide excellent results for well-behaved sets of measured data. Computing time required to obtain attached flows that are everywhere either laminar or tur- the results in Fig. 3 was 400 hours on a Cray-YMP pro- bulent. However, none can adequately model transition cessor. At the very low Mach number of 0.1 for this case, from laminar to turbulent flow, and few can satisfactorily the large difference between fluid velocity and speed of sound results in an extremely slow numerical integration handle separating and reattaching flows without empirical process. More recent work shows that for Mach numbers adjustments being made to the turbulence model. Example higher than about 0.4, computing time will only be about results obtained using the full Navier-Stokes, LES, and 25% of that required for this case.

RANS equations are presented later in this paper. Addi- tional details of these approximations can be found in Clearly, it still is not practical to solve the full Ref. 4.

Navier-Stokes equations for flows about arbitrary config- These equations are instrumental in the following urations at Reynolds numbers of most interest. Neverthe- less, a method now exists for providing benchmark results 13 research areas.

for all of the properties of flows undergoing transition, including heat transfer, and for arbitrary geometries.

Furthermore, this method can be used to search for ways to control the transition process. Having this computa- tional capability should greatly accelerate the

o Measured (Suder, et al.)

This advance, together with the ones discussed earlier 0 Measured (Sohn, St al.)

in connection with the prediction of boundary layer tran- - Computed (grid B) sition, has set the stage. For the first time, with the LES approximation, flows about aerodynamic shapes are cal- culated without use of pre-selected modeling "constants" 6 Turbulent correlation either for the transition process or for the fully-developed turbulence. As computers become more powerful, these advances could have profound implications for computing o4 aerodynamic characteristics of complete aircraft. It will be 000 00 possible to calculate laminar, transitional, and turbulent flows as well as attached, separating, and reattaching flows about arbitrary shapes. This will be done without C2 empiricism and with accuracy closely approaching that Cl) Blaslus (laminar) provided by the full Navier-Stokes equations. Flows about jet engine turbine and compressor blades at full-scale Reynolds numbers now can be attempted with today's 2.5 5.0 7.5 10.0 12.5 computers. In fact, such efforts already have begun.

Length Reynolds number x 10 Fig. 3 Measured and computed skin friction coefficient Turbines and Compressors along the length of a flat plate.

Remarkable advances in jet engine performance have search for better models to use in approximate predictive been made over the years, despite the complexity of com- methods for the transition process. Finally, this capability pact, multiple-stage compressors and turbines. Tradi- provides an approach for calculating flows about arbitrary tionally, the design process has relied heavily upon empir- geometries using either full Navier-Stokes equations or ical correlations developed from data bases provided by the more computationally efficient Large Eddy Simulation previous designs. Despite the maturity of engine technol- approximation.

ogy, considerable improvements are still possible, as shown by the study reported in Ref. 10. These improve- ments would be accelerated by availability of more gen- Turbulence Modeling eral and reliable predictive capabilities for time-dependent viscous flows through multiple-stage, axial-flow compres- Another major advance related to predicting viscous sors and turbines. Advanced designs that employ high compressible flows is the development of a new dynamic blade-loadings and small axial gaps between blade rows subgrid-scale model, reported in Ref. 9, for use with the preclude the use of previous methods of analysis that were LES approximation. This model appears to have over- based on isolated airfoil rows. Fortunately, new numerical come all deficiencies of the widely used Smagorinsky methods that treat multiple rows of rotor and stator blades model. The new model utilizes strain rate fields at two as a system are being developed rapidly.

different scales in the computed large-scale field to extrapolate the small-scale stresses in the unresolved One emerging method, based on the Reynolds- subgrid-scale region. This avoids the need to choose a dif- averaged Navier-Stokes equations, is an extension of ferent optimal constant for each flow being computed.

earlier work (presented in Ref. 11) in which the flow That is, the closure "constant" is determined as part of the through a single-stage turbine configuration was calculation of each flow and it is allowed to vary with simulated numerically. Initial results from this new local flow conditions. In addition, the model has the cor- method, for the flow through a 2-1/2-stage two- rect limiting behavior near walls: it vanishes in laminar dimensional compressor, are presented in Ref. 12.

flow, it properly represents dissipation and backscatter of Sample results from this work are presented in Fig. 4, energy from small scales to large scales in the laminar-to- where calculated time-averaged pressure coefficients for turbulent transition region, and it includes compressibility the second stage rotor and stator are compared with effects. Comparisons presented in Ref. 9 show excellent measurements made on a three-dimensional model.

agreement for various turbulence quantities (obtained Agreement between two-dimensional calculated results using the dynamic subgrid-scale model and the LES and midspan results from the three-dimensional test is equations) with solutions of the full Navier-Stokes quite good. This numerical work is being extended to equations.

three dimensions, and preliminary results for

2.0 Computation of flows about multi-element airfoils is 0 Measured a particularly challenging problem. In addition to geomet- Computed ric complexity, treatment of this problem requires con- 1.5 sideration of boundary layer separation, boundary layer and wake interaction, and Reynolds-number effects. A C very efficient method for treating these complexities was u 1.0 reported recently in Ref. 13. Results obtained by solving the incompressible Reynolds-averaged Navier-Stokes U equations, for both single and multi-element airfoils, are presented in the reference. An example of these numerical results, compared with measurements, is presented in Fig. 6 for a NASA 9.3% thick blunt-based supercritical airfoil at 14.25° angle of attack, with a leading edge slat deployed -47.2° and two trailing edge flaps deflected to ROTOR 30 and 49.7°, respectively. Chord Reynolds number is 2.83 million and the measured results were obtained at a -.5' Mach number of 0.201. Agreement between computed 0 2 4 8 and measured results is excellent, and the computation Inches from rotor leading edge required only about 4 mm on a single Cray-YMP proces- Fig. 4 Calculated and measured instantaneous surface sor. When methods similar to this one are extended to pressures on the blades of a 2-1/2-stage compressor. treat three-dimensional multi-element wings in the near future, it should be possible to accelerate development of instantaneous surface pressures on the blades of a new and less complicated high lift devices.

2-1/2-stage compressor are shown in Fig. 5 (see color page). A new high lift device now under study at the NASA Ames Research Center shows remarkable potential. Itis Clearly, these calculations are expensive to make similar to the Gurney Flap used in automobile racing, but with today's computers. However, the expected availabil- it has the distinguishing feature of being split so that it ity, before the turn of the next century, of at least an addi- may be located upstream of the airfoil trailing edge.

Unpublished computed and measured results for a 1.25% tional three orders of magnitude of machine processing chord flap located at the trailing edge of a NACA 4412 power will bring this technology into the hands of airfoil are shown in Fig. 7. Results are for a Reynolds designers.

High Lift Devices High lift devices are important features of both mili- -2°r1 tary and civil aircraft. They enable heavily loaded aircraft -1.5 o Measured to take off from runways of reasonable length and to Computed II maintain controlled flight during landing approaches at -1.0 reasonably low speeds. Their use also controls the size of the high-noise footprint in the vicinity of airports by enabling aircraft to gain altitude quickly. This latter capa- bility is particularly important for next-generation super- sonic transports that must be able to operate from existing • 51 I commercial airports without benefit of very high bypass 0 .5 1.0 1.5 2.0 ratio engines that are inherently quieter. Two recent tech- Fraction of chord nical advances will be presented to illustrate progress being made in this important area. The first is a very effi- Fig. 6 Surface pressure coefficients for a four-element cient and accurate computational method for predicting NASA 9.3% thick blunt-based supercritical airfoil. Lead- flows about multi-element airfoils, and the second is a ing edge slat at -47.2°; trailing edge flaps at 300 and very simple miniature split flap that has a remarkable 49.7°, respectively; Mach number = 0.201; angle of capability to increase airfoil lift.

attack = 14.25°.

VA

number of 4 million. The computed results were obtained High Angle of Attack Technology from an incompressible Navier-Stokes code, and the mea- sured results are for a very low subsonic Mach number.

NASA's High Alpha Technology Program is provid- The miniature split flap is quite effective in increasing the ing a greater understanding of the physics of massively lift throughout the angle of attack range, and it increases separated flows and vortex bursting, new tools for use in maximum lift of the airfoil by about 20%. The flap designing future highly maneuverable aircraft, and new produces a lift increment by reducing the extent of upper concepts for controlling aircraft at all possible attitudes.

surface trailing edge flow separation while increasing The program includes numerical analysis, wind tunnel downward deflection of the flow as it leaves the trailing experiments with both sub-scale and full-scale models, edge. Parametric studies are being performed to investi- flight simulator studies of new control laws for extreme attitude flight with thrust vectoring, and final validation gate different flap lengths and hinge-line locations.

using an F-18 testbed aircraft.

These examples provide an optimistic outlook for future advances in the development of high lift devices of Flow about a complete, complex geometry aircraft, at greater simplicity with attendant manufacturing and angles of attack where flow physics also are very com- maintenance cost reductions. plex, now can be calculated with sufficient precision to be of practical use in the design process. An example of this capability is shown in Fig. 8. There, pressure coefficients calculated by solving the Reynolds-averaged Navier- Stokes equations are compared with flight measurements 2.0 i- made at two stations on the F-18 high angle of attack research vehicle, flying at Mach number 0.243 and at 30.3° angle of attack. Inboard and outboard leading edge flaps are deflected 33° nose down, and the horizontal tail Flap deployed I is deflected nose down. Remarkable agreement between calculated and measured results is noted. Addi- 1.5 tional results, and a more complete discussion of the com- putational procedures, are presented in Ref. 14.

U Performing numerical calculations with the time- />/rected Computed lift Increment accurate version of the code provides time varying pres- 1.0_ ,, sure loads resulting from unsteady separated flow. Cou-

V

pling computed unsteady pressure loads with a structural o 0 Measured response code permits the study of buffet phenomena.

- - - - Computed lift Increment Having this capability is very important, since vertical tail due to flap buffeting is common on twin tail aircraft flying at high angles of attack and can be severe enough to cause 0 5 10 15 premature structural fatigue. Example results from Angle of attack (deg) Ref. 15, for one of the vertical tails on the F-18 experienc- ing buffeting, are presented in Fig. 9. Here the calculated Fig. 7 Effect of 1.25% chord miniature split flap on the buffet frequency is correlated with buffet frequencies variation of lift coefficient with angle of attack for a measured on two different size wind tunnel models and on NACA 4412 airfoil.

the F-18 in flight, It is noted that all of the results show reasonably good correlation.

Station 4 StatIon 7 1.5 Measured (flight) Computed 1.0 C 0 2 -1 -2 I -1.0 '• 0 45 90 135 0 .2 .4 .6 .8 .10 180 Azimuthal angle Fraction of body semlspan Fig. 8 Comparison of computed surface pressure coefficients with those measured in flight. Leading edge flaps 33° nose down; horizontal tail 7° nose down; Mach number = 0.243; angle of attack = 30.3°.

Flight Calculations of this type, that include both complex o Measured geometries and complex physics, require about 55 hr on a • Computed Cray-YMP processor for a single case. Improvements in Wind Tunnel numerical methodology, and availability of more power- 11%scaie ful computers, should bring this time down to the order of I i 0 Full scale minutes before the turn of the next century. Thus, applied aerodynamicists can expect to routinely use this capability 20I- for aircraft design in the not-too-distant future.

I U.

10I_

A

Powered Lift Computational tools available to assist designers of 0 50 100 150 200 powered lift vehicles have been quite limited. It is even Speed (m/sec) difficult to obtain from tests of scale models in wind tun- nels reliable quantitative aerodynamic performance data Fig. 9 Computed and measured F-I 8 vertical tail buffet for lift-off and hover in ground effect, and for transition frequency. Mach number = 0.24; angle of attack = 30°; from vertical to horizontal flight. This is because of Reynolds number based on chord = II x 106.

scaling effects sensitive to Reynolds number. On the other hand, full-scale testing of a large number of new concepts is not practical early in design cycles because of cost

considerations. Furthermore, previously available Aeroacoustics computational methods for inviscid flows, even with corrections for viscous boundary layers, have proved Commercial aircraft are facing increasingly strict inadequate. However, recent advances in supercomputer regulations governing noise in the vicinity of airports.

and algorithm capabilities now allow attempts to solve the Achieving jet noise suppression without incurring unac- three-dimensional Reynolds-averaged Navier-Stokes ceptable penalties in thrust loss, cost, and weight will be equations for flows about powered lift vehicles. an increasingly difficult challenge. Even for military air- craft, jet noise is becoming more of a concern. More pow- Efforts are under way to calculate the flow about a erful and more highly integrated engines are exacerbating complete AV-8B Harrier and to validate the results by airframe component fatigue due to jet noise, and previ- comparisons with data obtained from flights of a YAV-8B ously mentioned performance penalties result from the use of noise suppression devices. Lack of adequate meth- V/STOL research aircraft. Early results from this effort are presented in Ref. 16. An example of the complexity of ods for predicting source noise, and its propagation to the the flow surrounding the Harrier flying at 80 angle of far field, have hampered progress in finding solutions to attack 30 feet above the ground, with a forward Mach these problems. Recent important advances have been number of 0.04 is shown in Fig. 10 (see color page). made, however, in the development of computational There, calculated instantaneous streamlines are shown for methods for predicting sound pressure levels in both the only a portion of the entire flowfield. In addition to pro- near and far fields. These offer the hope of providing new viding instantaneous aerodynamic loads, lift, drag, pitch- analysis tools to assist in developing and evaluating sup- ing moments, and knowledge of whether or not hot gases pression concepts.

are being ingested into the inlets, the computational method also can provide valuable information about The results of a study presented in Ref. 17 show one "suckdown" in ground effect. This latter information can advance in which the far field sound from a pair of com- be obtained by performing calculations with and without pressible co-rotating vortices is calculated. Sound pressure engines operating and then comparing the results. In this levels obtained from solutions of the full Navier-Stokes example, without the wing flaps deflected, calculations equations were found to agree within 3% with estimates predicted a net loss of about 17%of the engine thrust due made using acoustic theory for this two-dimensional to "suckdown" effects. problem. This result is very impressive. However, integration of the full three-dimensional Navier-Stokes Dedicated flight tests of the YAV-8B are planned to equations into the far field for high Reynolds number validate this new computational capability, and one exper- flows and complex surface boundaries would involve iment already has been performed. In it, the possibility excessive computational cost at this time.

was investigated of using infrared imagery as a nonintrusive method for providing qualitative data for jet Results of another recent study reported in Ref. 18 trajectories and major flow structures around the vehicle. offer a possible approach for making computational An example of initial results taken from Ref. 16 is shown aeroacoustics affordable sooner. This optimism is in Fig. 11 (see color page). Infrared measurements made predicated on development of an accurate finite difference during a flight of the YAV-8B are compared with method for calculating acoustic waves. This would pro- calculations. Note that outflow from the rear jets only may vide a unifying approach for connecting near field noise be visualized with this technique since the temperature of source predictions made with finite difference methods fan air flow emanating from the front jets is not hot with the acoustic field equations governing the far field. A enough to produce an infrared image. Further work is fourth order finite difference algorithm that maintains planned to quantify information provided by this integrity of phase correlated waves over long distances is technique.

derived in the reference work. In addition, the method allows waves to freely radiate beyond mesh boundaries.

Calculation of flows about powered lift vehicles is Essentially first principle predictions of aircraft jet noise computationally expensive at this time. Results discussed in the far field should now be possible. Inner boundary above now require about 40 hr of calculation on a Cray- conditions for this acoustic propagation method would YMP processor. Of course, this time is expected to be need to be obtained from time-accurate near field Navier- reduced to a matter of minutes before the end of the Stokes solutions like the method described above for the decade through improvements in algorithms and comput- source noise. Having such a computational approach ers. At that time, applied aerodynamicists can use this should hasten development of improved noise suppression new capability to greatly improve performance and effi- techniques.

ciency of powered lift vehicles.

Rotorcraft An initial attempt at this problem is presented in Ref. 20. The two-dimensional download flow field is cal- Aerodynamic flows about rotorcräft are extremely culated by solving the velocity-vorticity formulation using complicated. These flows, even for low speed or hover a staggered-grid approach in a fully coupled manner. In conditions, are mathematically nonlinear, three- this method, vorticity is defined at mesh nodes, and veloc- dimensional, and unsteady, and have regions of transonic ity components are defined at mesh-cell sides. This flow near blade tips. Wakes shed from rotating blades are arrangement allows for accurate representation of the complex vortical flows that may interact with other blades definition of vorticity at node points and for conservation and the fuselage. All of these complications have made it of mass at cell centers. Additionally, the fully coupled very difficult to develop suitable predictive methods for method provides for use of an implicit method that use in design of rotorcraft.

requires only a single iteration to converge at each time step. A comparison of calculated and measured pressure Until recently, application of computational fluid distributions on the XV-15 airfoil is shown in Fig. 12. The dynamics methods has been limited to isolated compo- negative pressure peak is accurately predicted in location nents of complete vehicles. However, aerodynamic inter- but underpredicted in absolute value. Calculated lower- actions between rotating and nonrotating components are surface pressures are in excellent agreement with mea- widely recognized as major challenges that must be sured results.

addressed before computations can be a major contributor .4 to the design process. This is discussed more fully in Ref. 19. Measured progress has been slow, consisting -3 mostly of conceptualization and development of viable strategies for interfacing different computational grid -2 structures. One such method is described in Ref. 19. It defines a computational domain about an arbitrary body in C which the rotor is represented by an actuator disk. A oc Navier-Stokes technique that admits a prescribed pressure jump across the actuator disk is then applied. This simpli- 0 fied representation of rotor blades will be replaced by a more accurate finite-difference simulation when comput- ing power permits. In the meantime, this method can be a.

used to study effects of the rotor wake on nearby fuselage or wing components.

Tiltrotor designs offer many opportunities for increased rotorcraft performance for both civil and mili- I I I I I I tary applications. Tiltrotor aircraft can take off and land 0 .2 .4 .6 .8 1.0 vertically, and fly at more than twice the speed of con- FractIon of chord ventional helicopters. In the civil commuter application, vertiports can be placed either at downtown metropolitan Fig. 12 Comparison of calculated and measured average locations or at remote small towns. Passengers can be surface pressure coefficients on the XV-15 airfoil; flap transported at high speed either to downtown locations in deflection = 600.

other cities several hundred miles away, or to hub airports where these aircraft would not have to compete for use of Although computational methods for predicting already overcrowded main airport runways.

rotorcraft aerodynamics have progressed significantly in recent years, an accurate finite-difference simulation of A significant challenge in the design of tiltrotor con- complete flow fields about these vehicles still is not feasi- figurations results from the wing being immersed in the ble. Limitations exist in many areas including computer rotor wake during hover. This produces a download on the power, algorithms, grid generation, turbulence models, wing which reduces its lifting effectiveness during take- and wake simulations. However, the technical challenges off and landing. The result is a reduction in payload that are being addressed and eventually will be met.

can be carried. Complete analysis of this situation requires calculation of complex three-dimensional flows about a wing at 90° incidence.

lift-to-drag ratio is considerably greater than that for the Oblique Wing delta wing throughout the Mach number range and even Substantial improvements in both performance and exceeds the lift-to-drag ratio for a conventional transport efficiency can be realized by abandoning the notion that at high subsonic speeds.

aircraft must possess bilateral symmetry. A number of years ago, R. T. Jones presented analyses in Refs. 21 and Considerable additional work still is required to 22 showing the optimum wing, for both supersonic and resolve human factors issues associated with asymmetric high subsonic Mach numbers, to be one that pivoted about aircraft, handling qualities and control laws, engine instal- its centerline with one tip pointing forward and the other lation issues, landing gear arrangements, and a number of tip pointing aft. The pivot angle, or sweep, can be varied other technological uncertainties. Nevertheless, there are with Mach number such that the component of Mach no known technology roadblocks to the development of number normal to the leading edge always is as close to an oblique all-wing aircraft.

the optimum as possible. Since the theories were pub- lished, aerodynamic advantages have been validated by wind tunnel tests of various wing-body combinations Hypersonics having obliquely swept wings. Feasibility also has been demonstrated in flights of both manned and unmanned Advent of the National Aerospace Plane Program aircraft. (NASP) stimulated renewed interest in development of technology for hypersonic flight. The goal of the NASP is Interest in development of a second-generation super. to demonstrate, with the X-30 research vehicle, the capa- sonic transport has prompted new studies of the oblique bility to take off from a conventional runway and fly to wing concept. Now flying wings are no longer an oddity, nearly orbital speeds using airbreathing propulsion. This electronic stabilization of aircraft is a reality, and very is an extremely challenging task since there are no large aircraft are required to satisfy traffic demand. So an ground-based test facilities that can closely, match the oblique all-wing aircraft is more attractive than ever complete flight environment at Mach numbers above before. The oblique flying wing theoretically has maxi- about 8.

mum aerodynamic performance over a broad range of Mach numbers (both subsonic and supersonic). In addi- The last aircraft-like vehicle that this country devel- tion, it has theoretical maximum structural efficiency, can oped for flight from orbital to landing speeds is the Space easily meet the FAR-36 Stage Ill noise standards and Shuttle orbiter. That task was somewhat simpler than the (studies show) it might be able to perform the mission of a one facing the developers of the X-30 since the orbiter is 747-400 aircraft at twice the speed and about the same boosted to orbit by rocket engines and is unpowered dur- cost. Aerodynamic efficiency is illustrated in Fig. 13. The ing reentry. Nevertheless, it was very difficult, and mea- estimated variation of lift-to-drag ratio with Mach number surements made during orbiter flights have shown where improvements are needed in tools for designing hyper- of an oblique all-wing transport is compared to that for a more conventional delta wing design. The oblique wing sonic vehicles.

One such area needing improvement is that involving the determination of aerodynamic characteristics of vehi- b Oblique all wing cles flying at very high speeds where the airflow about the

I

vehicle does not behave as a perfect gas. This is illustrated

g 15 I

-

I by results presented in Fig. H, in which the predicted I 0.

., center of pressure location on the orbiter is compared with transport I measurements made during pullup/pushover maneuvers I on the flight of STS-2 at Mach number 21. These results were obtained from Ref. 23. Note that preflight predic- tions based on wind tunnel data determined the center of 0 .5 1.0 1.5 2.0 .25 pressure to be between 0.6 and 0.8% of the orbiter body Mach number length more aft than the location measured in flight. For- tunately, sufficient trim power provided by the body flap Fig. 13 Calculated variations of lift-to-drag ratio with was available to compensate for this large unexpected dif- Mach number for a delta wing supersonic transport, an ference between predicted and actual center of pressure oblique all-wing transport, and a 747-400 transport.

locations. Subsequent analysis attributed this large

.680 STS-2 Mach number 21 = Airfoil, chord 10 m C = V 7 km/s. altitude = 74 krn, angle-of-attack = 40° .!

.676 'I,

cIII

.

1.2 .672 / a • .668 e

0I

a 0.

! .8 N a .664 C 0.

C-) L.

C C .660 .664 .668 .672 .676 .680 C ' .4 Center of pressure, fraction of body length C) -j Fig. 14 Comparison of shuttle orbiter longitudinal center C.)

of pressure measured in flight, with prcdictions bascd on wind tunnel data for pullup/pushover maneuvers.

difference to real gas effects which were not adequately Center-of- CL CD pressure shift known when the vehicle was designed.

(real gas-perfect gas) Improvements in understanding how to account for Fig. 15 Comparison of calculated real gas and perfect air real gas effects have been made since the design of the airfoil lift and drag coefficients and center of pressure orbiter. In addition, today's more powerful computers shift.

permit calculations including these complex physics to be made with a reasonable amount of computer time. An complete hypersonic vehicles, including flows through example of effects of real gas environments on aerody- scramjet engines with additional complexities of namic characteristics of an airfoil, with a chord length combustion.

corresponding to that at the mid semispan location of the shuttle orbiter wing, is shown in Fig. 15. Real gas reacting flow chemistry was calculated with a code that accounts Measurement Techniques for both thermal and chemical nonequilibria in the shock layer. Also shown for comparison are results calculated Numerous advances are being made both in measur- by assuming perfect gas behavior (constant ratio of spe- ing various flow quantities without interference by the cific heats). Lift and drag coefficients are not strongly sensors and in more closely simulating flight environ- affected by real gas considerations. However, center of ments in ground-based facilities. One such advance pressure location from the real gas calculation is more reported in Ref. 25 is the development of a method for than 1% of chord forward of the location based on the nonintrusively obtaining pressures everywhere on an perfect gas estimate. This forward shift is consistent with aerodynamic surface at any given instant of time. In this that found in the studies of shuttle orbiter aerodynamics.

method, the surface is coated with an oxygen permeable These results are discussed in more detail in Ref. 24.

polymer containing a luminescent molecule and then illuminated with ultraviolet radiation. The luminescence Accurate computational methods for hypersonic intensity distribution, which is related to the pressure, is flows are a necessity because of the lack of ground-based obtained using quantitative video techniques while the test facilities capable of reproducing real gas environ- surface is subjected to a flow. These results are then con- ments. Applied aerodynamicists must rely extensively on verted to a surface pressure map with the aid of computer the computer to design the X-30 vehicle. Accordingly, processing. An example of early results taken from considerable emphasis is currently being placed on devel- Ref. 25 is given in Fig. 16. Upper surface pressure oping and validating methods for calculating flows about

-3 method, described in Ref. 27, is a real-time interferometry system capable of providing densities and pressures in both steady and unsteady flows. Pressure distributions near the leading edge of an oscillating airfoil in compress- -2 ible flow that were obtained using this method are presented in Ref. 28. Data such as these, together with C velocity components measured with laser velocimetry, will be invaluable for improving understanding of aero- -1 dynamic flows and for validating computational methods.

Important advances also are being made in providing wind tunnel test environments more representative of free-air flight. These advances will permit studies of boundary layer transition and methods for its control in ground-based test facilities. One approach to obtaining low disturbance "quiet" flow in a small supersonic wind I I tunnel is reported in Ref. 29. This, and other approaches 0 .2 .4 .6 .8 1.0 soon will be ready for use in large production-type Fraction of chard facilities.

Fig. 16 Upper surface pressure coefficients at the midspan of a two-dimensional wing with an NACA-0012 Numerical Optimization section. Mach number = 0.66; angle of attack = 0.

The shape of an airplane is influenced by many fac- coefficients at the midspan of a two-dimensional wing tors other than aerodynamics. Among these are require- with an NACA-0012 airfoil section are presented for a ments imposed by desired dimensions of the payload flow Mach number of 0.66. Also shown for comparison compartment(s), mission range which influences fuel tank are midspan chordwise pressures measured with volume, dimensions of available engines, environmental conventional pressure taps. Excellent agreement between considerations (sonic boom), and observability (military).

the two types of measurements is noted.

Finding the optimum aerodynamic shape that satisfies constraints imposed by requirements such as these is a Reaction of the paint to pressure changes is reversible formidable task. It usually involves sorting through many and rapid. This means that the same coating can be used possible configurations, suggested either by theoretical over and over again, and likely can be used for pressure analysis or by experimentation, until the one having the measurements in unsteady flow situations.

"best" performance is obtained.

More recently, this method has been applied in wind Availability of efficient numerical methods for calcu- tunnel tests of an oblique wing at supersonic speeds, and lating aerodynamic performance made it feasible to cou- results are presented in Ref. 26. In addition, an initial ple numerical optimizers to aerodynamic analysis codes to study of the use of luminescent pressure sensitive paint in find an optimum shape based on various imposed con- the flight environment using an F-104 airplane has been straints. Because of computer limitations, most applica- conducted and will be reported. Promising flight-test tions of numerical optimization have been limited either results were obtained along with information needed to to two-dimensional airfoils or to use of the linearized flow improve application of the method in flight. Also under equations. Computers are now powerful enough, and study is an extension of the method to simultaneously analysis codes based on the nonlinear flow equations are measure both temperature and pressure. Within the next now fast enough, to apply numerical optimization to real- few years, applied aerodynamicists should have a new istic problems. - tool that will provide pressures and temperatures every- where on a test model, for both steady and unsteady flow An example of the numerical optimization of a high- conditions, all without installation of sensors.

speed civil transport wing by using the Euler equations is presented in Ref. 30. The objectives were to start with a Also under development arc various methods for given wing planform and volume and find the airfoil nonintrusively determining off-surface field measure- shapes along the span that would provide maximum lift- ments of temperature, density and pressure. One such to-drag ratio at a Mach number of 2.1, while keeping the

wing leading edge radii and volume constant. Results for goal of the HPCCP to advance scientific computational capabilities by a factor of 1000 before the end of the both baseline and optimized wings are shown in Fig. 17.

Use of the optimizer improved lift-to-drag ratio at the decade. Without increases in available computer speed of design Mach number by 5.3%. It is interesting to note that this order, the ability to perform multidisciplinary analy- lift-to-drag ratio at an off-design Mach number of 0.8 is ses would remain limited. Additional background and dis- about 20% higher than for the baseline wing. Obtaining cussion of goals of the HPCCP is presented in Ref. 31.

these results required about 5-1/2 hr (340 mm) of CPU time on a single Cray-YMP processor. This computing Aeroelasticity is an example of multidisciplinary phenomena that has been treated computationally for time should be reduced to less than 2 hr by next- many years. This was made possible with already avail- generation supercomputers just coming to the able computers by using forms of the governing equations marketplace. Before the end of this decade, applied aerodynamicists should be able to optimize an entire that approximate underlying physics. With today's com- puters, it is practical in a research environment to couple airplane shape by using Reynolds-averaged Navier-Stokes the Reynolds-averaged Navier-Stokes equations with any equations together with an optimizer.

one of a number of methods for approximating aeroelastic behavior of an aircraft structure. With future computers, one can envision coupling a NASTRAN code to a fluid dynamics code to provide even more precise results.

Another example of multidisciplinary analysis is the calculation of real gas phenomena including reacting flow chemistry associated with very high speed flight, hot exhaust flows interacting with aerodynamic surfaces, and combustion in the presence of flow through an engine.

Still another is prediction of electromagnetic signatures where interactions are modeled of incident radiation with materials and irregular surface shapes.

0 These initial attempts to solve various multidisci- plinary problems are pioneering methodology that even- tually will be used to couple all disciplines important to aircraft design such as aerodynamics, structures, propul- sion, and active controls. It will even be possible to include the discipline of electromagnetics when required.

Currently available computer power limits the num- ber of disciplines that can be treated simultaneously, even when approximate forms of the governing equations are used. However, as computer power increases, more disci- 0 .05 .10 .15 plines and less restrictive forms of governing equations Lift coefficient can be introduced. Eventually, multidisciplinary design codes will be available that will provide solutions to prob- Fig. 17. Comparison of lift-to-drag ratios for basclinc and lems with the use of reasonable amounts of computer optimized wings. Mach numbcr = 2.1.

time. Then, the time required to develop a new aircraft will be greatly reduced and improvements in the various measures of aircraft performance will be greatly Multidisciplinary Analysis increased.

Initiation of the High Performance Computing and Communication Program (HPCCP) has stimulated addi- Concluding Remarks tional interest in developing computational tools capable of treating strongly coupled multidisciplinary phenomena, Aerospace is important to the country's economic such as aerodynamics, structures, propulsion systems, and competitiveness and military superiority. Both continued active controls. This interest was stimulated, in part, by a erosion of the U.S. world market share for commercial

Phys. Fluids A, Vol. 3, aircraft while the demand worldwide for aircraft is grow- Turbulence and Scalar Transport," ing, and the concurrent reduction in expenditures for 1991, pp 2746-2757.

military aircraft, give applied aerodynamicists an unprecedented challenge to provide technology to offset 10Koff, B. L, "Spanning the Globe with Jet these trends. A number of emerging research results have Propulsion," AIAA Paper 91-2987, Arlington, VA, 1991.

been discussed that will assist applied aerodynamicists in meeting this challenge in the relatively near term. How- Rai, M. M. and Madavan, N. K., "Multi-Airfoil ever, substantial investments must continue to be made in Navier-Stokes Simulations of Turbine-Rotor Interaction," basic and applied research efforts, focusing on increasing AIAA Paper 88-0361, Reno, NV, 1988.

understanding of fluid dynamic phenomena, identifying new aerodynamic concepts, and providing validated tech- 12Gundy-Burlet, K. L, Rai, M. M., Stauter, R. C., nology for new aircraft, in order to continue to meet the and Dring, R. P., "Temporally and Spatially Resolved Flow in a Two-Stage Axial Compressor, Part 2– challenge in the more distant future.

Computational Assessment," ASME Paper 90-GT-299, International Gas Turbine Conference, Brussels, Belgium, References June 1990.

1 Anonymous, "1991 Year-End Review and 13 Rogers, S. E., Wiltberger, N. L., and Kwak, D., Forecast—An Analysis," Aerospace Industries "Efficient Simulation of Incompressible Viscous Flow Association, Washington, DC, 1991. Over Single and Multi-Element Airfoils," AIAA Paper 92-0405, Reno, NV, 1992.

2Anonymous, "1992 Current Market Outlook," Boeing Commercial Airplane Group, Seattle, WA, 1992. 14Rjzk, Y. M. and Gee, K., "Numerical Prediction of the Unsteady Flowfield Around the F-18 Aircraft at Large 3Peterson, V. L, Kim, J., Hoist, T. L., Deiwert, G. S., Incidence," AIAA Paper 91-0020, Reno, NV, 1991.

Cooper, D. M., Watson, A. B., and Bailey, F. R., 15 Rizk, Y. M., Guruswamy, G., and Gee, K., "Supercomputer Requirements for Selected Disciplines Vol. 77, No. 7, July "Numerical Investigation of the Tail Buffet on the F-18 Important to Aerospace," Proc. IEEE, Aircraft," AIAA Paper 92-2673, Palo Alto, CA, 1992.

1989.

16Smith, M., Chawla, K., and Van Dalsem, W., 4Peterson,V. L., "Impact of Computers on Aerodynamics Research and Development," Proc. IEEE, "Numerical Simulation of a Complete STOVL Aircraft in Vol. 72, Jan. 1984. Ground Effect," AIAA Paper 91-3293, Baltimore, MD, 1991.

5 Rai, M. and Mom, P., "Direct Numerical Simulation of Transition and Turbulence in a Spatially Evolving 17Mitchell, B. E., Lele, S. K., and Mom, P., "Direct Boundary Layer," AIAA Paper 91 -1607, Honolulu, Hi, Computation of the Sound from a Compressible Co-Rotating Vortex Pair," AIAA Paper 92-0374, Reno, 1991.

NV, 1992.

6Suder, K. L., O'Brien, J. E., and Reshotko, E., ' 8Davis, S., "Low-Dispersion Finite Difference "Experimental Study of Bypass Transition in a Boundary Layer," NASA TM-100913, 1988. Methods for Acoustic Waves in a Pipe,"J. Acoust. Soc.

Am., Vol. 90, No.5, 1991, pp. 2775-2781.

7Sohn, K. H. and Reshotko, E., "Experimental Study of Boundary Layer Transition with Elevated Free-Stream 19McCroskey, W. J., "Some Recent Applications of Turbulence on a Heated Flat Plate," NASA CR-187068, Navier-Stokes Codes to Rotorcraft," Fifth Symposium on 1991. Numerical and Physical Aspects ofAerodynamic Flows, Long Beach, CA, Jan. 13-16, 1992.

8White, F. M., "Viscous Fluid Flow," McGraw-Hill Book Company, 1974. 20Stremel, P. M., "Calculation of Unsteady Airfoil Flows with and without Flap Deflection at –90 Degrees 9 Moin, P., Squires, K., Cabot, W., and Lee, S., "A Incidence," AIAA Paper 91-3336, 9th Applied Dynamic Subgrid-Scale Model for Compressible Aerodynamics Conference, Baltimore, MD, Sept. 23-26, 1991.

21iones, R. T., "Theoretical Determination of the 27Brock, N. J., Chandrasekhara, M. S., and Carr, Minimum Drag of Airfoils at Supersonic Speeds," L W., "A Real Time Interfemmetry System for Unsteady J. Aerospace Sciences, Vol. 19, No. 12, Dec. 1952, Flow Measurements," ICIASF '91 RECORD, Rockville, pp. 813-822. MD, Oct. 25-28, 1991.

22iones R. T., "Aerodynamic Design for Supersonic 28Carr, L W., Chandrasekhara, M. S., and Brock, Speeds," Proceedings of the First International Congress N. 3., "A Quantitative Study of Unsteady Compressible in the Aeronautical Sciences, Madrid, Sept. 1958, Flow on an Oscillating Airfoil," AIAA Paper 91-1683, Advances in the Aeronautical Sciences, Pergamon Press, Honolulu, HI, 1991.

N.Y., 1959, pp. 34-51.

29Wolf, W. D., Laub, J. A., King, L. S., and Reda, 23Arrington, J. P. and Jones, J. J., "Shuttle D. C., "Development of the NASA-Ames Low- Disturbance Supersonic Wind Tunnel for Transition Performance: Lessons Learned," NASA CP-2283, Hampton, VA, 1983. Research up to Mach 2.5," AIAA Paper 92-3909, Nashville, TN, July 6-8, 1992.

24Park, C. and Yoon, S., "Calculation of Real-Gas Effects on Airfoil Aerodynamic Characteristics," AIAA 30Chang, I-C and Torres, F. 3., "Wing Design Code Paper 90-1712, Seattle, WA, 1990. Using Three-Dimensional Euler Equations and Optimization," AIAA Paper 91 -3190, Baltimore, MD, 25McLachlan, B. 0., Kavandi, 1. L., Callis, J. B., 1991.

Gouterman, M., Green, E., Khalil, 0., and Burns, D., "Surface Pressure Field Mapping Using Luminescent 31 Holst, T. L., Salas, M. D. and Claus, R. W., "The Coatings," accepted for publication in Experiments in NASA Computational Aerosciences Program—Toward Fluids. Teraflops Computing," AIAA Paper 92-0058, Reno, NV, 1992.

26McLachlan, B. G., Bell, J. H., Kennelly, R. A., Schreiner, 3. A., and Strong, 3. M., "Pressure Sensitive Paint Use in the Supersonic High-Sweep Oblique Wing (SHOW) Test," AIAA Paper 92-2686, Palo Alto, CA, 1992.

'r - ç r I inc crngcc urtace pressule coefticient in tile I ic Fig. 10 Calculated tiow about a Harrier aircraft 30 feet second stage of a 2-1/2-stage compressor.

above the ground. Mach number = 0.04; angle of attack = 8°.

Fig. 11 Flow about a Harrier aircraft. (a) Computed, (b) Infrared measurements.

PtE B! AN( NOT FILMED PPE€D

REPORT DOCUMENTATION PAGE Public reporting burden for this collection of information Is estimated to average 1 hour per response, Including the time for reviewing instructions, searching existing data sources, gathering and maintaining the data needed, and completing and reviewing the collection of Information. Send commentS regarding this burden estimate or any other aspect of this collection of Information, including suggestions for reducing this burden, to Washington Headquarters Services, Directorate for Information Operations and Reports, 1215 Jefferson Davis Highway, Suite 1204. Arlington, VA 22202-4302. and to the Office of Management and Budget Paperwork Reduction Proiect (0704-0188). Washington, DC 20503.

(Leave blank) 2. REPORT DATE 1. AGENCY USE ONLY 3. REPORT TYPE AND DATES COVERED February 1993 Technical Memorandum 4. TITLE AND SUBTITLE 5. FUNDING NUMBERS Applied Aerodynamics: Challenges and Expectations 505-90 6. AUTHOR(S) Victor L. Peterson and Charles A. Smith 8. PERFORMING ORGANIZATION 7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) REPORT NUMBER Ames Research Center Moffett Field, CA 94035-1000 A-92160 SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSORING/MONITORING 9.

AGENCY REPORT NUMBER National Aeronautics and Space Administration NASA TM403963 Washington, DC 20546-0001 11. SUPPLEMENTARY NOTES Point of Contact: Charles A. Smith, Ames Research Center, MS D-200-1, Moffett Field, CA 94035-1000; (415) 604-5113 Presented at AIAA 10th Applied Aerodynamics Conference, June 22-24, 1992, Palo Alto, CA.

128. DISTRIBUTION/AVAILABILITY STATEMENT 12b. DISTRIBUTION CODE Unclassified - Unlimited Subject Category 02 13. ABSTRACT (Maximum 200 words) Aerospace is the leading positive contributor to this country's balance of trade, derived largely from the sale of U.S. commercial aircraft around the world. This powerfully favorable economic situation is being threatened in two ways. (1) The U.S. portion of the commercial transport market is decreasing, even though the worldwide market is projected to increase substantially. (2) Expenditures are decreasing for military aircraft, which often serve as proving grounds for advanced aircraft technology. To retain a major share of the world market for commercial aircraft and continue to provide military aircraft with unsurpassed performance, the U.S. aerospace industry faces many technological challenges.

The field of applied aerodynamics is necessarily a major contributor to efforts aimed at meeting these technological challenges. A number of emerging research results that will provide new opportunities for applied aerodynamicists are discussed in this paper. Some of these have great potential for maintaining the high value of contributions from applied aerodynamics in the relatively near future. Over time, however, the value of these contributions will diminish greatly unless substantial investments continue to be made in basic and applied research efforts. The focus: to increase understanding of fluid dynamic phenomena, identify new aerodynamic concepts, and provide validated advanced technology for future aircraft.

15. NUMBER OF PAGES 14. SUBJECT TERMS Applied aerodynamics 16. PRICE CODE A02 17. SECURITY CLASSIFICATION 18. SECURITY CLASSIFICATION 19. SECURITY CLASSIFICATION 20. LIMITATION OF ABSTRACT OFTHIS PAGE OF ABSTRACT OF REPORT Unclassified Unclassified NSN 7540-01-280.5500 Standard Form 298 (Rev. 2-89) Prescribed by ANSI Std. Z39-18 298.102