V/STOL wind-tunnel testing

NASA Technical Memorandum 85936

V/STOL Wind-Tunnel Testing

David G. Koenig, Ames Research Center, Moffett Field, California

N/ SA

National Aeronautics and

Space Administration

Ames Research Center

Moffett Field, California 94035

V/STOL WIND-TUNNEL TESTING David G. Koenig NASA Ames Research Center_ Moffett Field, California 94035, U.S.A.

SUMMARY Factors influencing effective program planning for V/STOL wind-tunnel testing are discussed. The plan- ning sequence itself, which includes a short checklist of considerations that could enhance the value of the tests, is also described. Each of the considerations, choice of wind tunnel, type of model installation, model development and test operations is discussed, and examples of appropriate past and current V/STOL test programs are provided. A short survey of the moderate to large subsonic wind tunnels is followed by a review of several model installations, from two-dimensional to large-scale models of complete aircraft configurations.

Model sizing, power simulation, and planning are treated, including three areas in test operations: data- acquisition systems, acoustic measurements in wind tunnels, and flow surveying.

SYMBOLS A diameter of lifting element, m (ft) aspect ratio, b2/S, nozzle or jet area, DL m2 (in. 2 ) diameter of nozzle, m (ft) lifting element area [n(_/4)DL2], m2 (ft 2) Dn A L outside diameter of annular nozzle slot (the momentum area of aircraft (_b2/4), m2 (ft 2) Do A M ° investigation of Ref. 35), m (ft) A N force parallel to free stream, N (Ib) area of powered lift element, fully Fx expanded; or equivalent nozzle area, m2 (ft 2 ) h ground height, m (ft) cross-sectional area of wind-tunnel test H test section height, m (ft) AT section, m2 (ft _) tail incidence, deg b wing span, m (ft) it IGE in ground effect B test section width, m (ft) K angle-of-attack correction factor used in c wing chord, cm or m (in. or ft) Ref. 52, [3.537 - 1.256(1 - M2)I/2 mean aerodynamic chord, m (ft) 1 length, m (ft) section normal force coefficient L lift, N (Ib) c n mass rate of flow of jet, kg/sec (Ib/m/sec) pressure coefficient, (p - p_)/q mj Cp corrected coefficients M Mach number Ccorr CD drag Coefficient, total if not otherwise n number of lifting elements indicated, drag/(qS) NPR nozzle pressure ratio aerodynamic drag coefficient, C D (total) CDA less effects of thrust and ram drag OGE out-of-ground effect jet thrust/qS OASPL overall sound pressure level Cj lift coefficient, total, (L/q)S p local static pressure, N/m 2 (Ib/ft 2) CL total pressure, N/m 2 (Ib/ft 2) CLA aerodynamic lift coefficient C L (total) Pt less effects of thrust P_ free-stream static, N/m 2 (Ib/ft 2) CLAERO same as CLA Po free-stream total pressure, N/m 2 (Ib/ft 2) total moment coefficient, moment/q S_ C m PNL perceived noise level aerodynamic moment coefficient less Cm A effects of thrust and ram drag q dynamic pressure, free stream unless noted otherwise, I/2 pV 2, N/m 2 (Ib/ft 2) measured coefficients Cmeas s distance along jet, m (in.)

see Fig. 71 Csi S wing reference area, m2 (ft 2) C u jet momentum coefficient (mj/q). (Vj/S) SPL sound pressure level d jet exit diameter (_Do) T static thrust, N (Ib) D equivalent surface diameter of the suck- down plate, m (ft) Tc thrust coefficient, [thrust/(q × fan area)] moving belt speed, m/sec (ft/sec) Be equivalent exit diameter, diameter of a V B circle whose area equals the sum of the areas of all jets, m (ft) V e effective velocity ratio, [qo/(T/2AN)] I/2 _v deflector vane setting withrespect to fan

axis,deg

I Ve value of V e at which the jet impinges on the floor, (see Fi_. 66) r vortex circulation strength Vj jet velocity, m/sec (ft/sec) aj jet deflection, deg Vo free-stream velocity, m/sec (ft/sec) _ difference or change in a parameter Wp primary weight rate of flow, kg/sec A wing sweep, deg ( 1 b. m/sec) p mass density of air w s secondary weight rate of flow, kg/sec (Ib.m/sec) Subscripts: x streamwise or longitudinal dimension, c corrected m (ft) e effective y spanwise dimension, m (ft) i inlet z vertical dimension, m (ft) n nozzle angle of attack, deg o free-stream conditions _g geometric angle of attack, deg u uncorrected 1. INTRODUCTION Wind-tunnel testing continues to be a key activity in aircraft development. It has been particularly necessary in the development of V/STOL aircraft because our understanding of the complex flow patterns affecting a powered-lifting system and formulations of prediction theories, has continued to lag the require- ment to design and build a V/STOL aircraft. Our present ability to predict details of aircraft performance and of stability and control relies heavily on the use of integral flow modeling combined with paneling methods based on potential-flow assumptions. The solutions and computer codes that apply the methods must be efficient enough to be used with aircraft design optimization techniques. As computer capability improves, computational fluid dynamics (CFD) will be used.

In any case, wind-tunnel testing will figure in many stages of the aircraft development from the initial conception of the thrust-vectoring method using quasi-two-dimensional or semispan installations to testing of full or scaled models of the aircraft configuration. Even though we do not have complete and reliable compu- tation methods with which to supplant wind-tunnel testing, they will, in whatever stage of development, be an integral part of the testing phase as means of correlating test results and, possibly, helping to restrict the size of the test matrix.

As indicated by the other lecturers in this series, and in the overviews on V/STOL concepts given in Refs. 1-3, it is a continuing objective in V/STOL development to make efficient use of all available on-board energy for meeting requirements in maneuverability, maximum speed, range, and short field lengths (at an affordable cost). The techniques in testing must be chosen carefully if it is to be ensured that the aircraft designer has well-documented test data (which verify predictions) providing him with sufficient information for preliminary design of the advanced version of the aircraft. At the same time, it is hoped that all the test data can be sufficiently documented using the proper instrumentation to make it valuable as a data base to be used for general aerodynamic study.

Because of these considerations, this lecture will concentrate on methodology with which an experimental- ist must be intimately familiar in order to plan and complete a successful V/STOL wind-tunnel test program.

The approach draws heavily on some of the factors mentioned in Refs. 4 and 5, and it is assumed that the reader can become familiar with the many other papers on V/STOL wind-tunnel testing, many of which will be cited throughout the lecture. It will also be assumed that the basics of Pope and Harper (Ref. 6) and of Pope and Goin (Ref. 7) are available. The lecture is organized into the following sections: _: an introductory section outlining the development of a program; wind tunnels: a brief review of typical, moderate-to-large size, testing facilities; V/STOL testing installations: a discussion of options or cate- gories of wind-tunnel installations; model development: a consideration of model sizing criteria, possible power systems, and construction techniques; test operations: a discussion of data acquisition systems, acous- tic testing, and flow visualization.

Treatment is limited primarily to fixed-wing (V/STOL) testing as opposed to rotor-wing (helicopter) test- ing. The author assumes V/STOL also applies to the tilt rotor or even to such designs as the X-wing (which will not be mentioned), but there will be no treatment of rotary-wing testing.

2. PLANNING AND PRETEST PREPARATION Before all the available tools for V/STOL wind-tunnel testing are evaluated in detail, the essential objectives of the program must be considered, and, to a certain extent, the data analysis be addressed.

Unfortunately, such factors as budgets, time restrictions, continual technological advancements, and military/ commercial objectives prohibit writing this part in a textbook fashion, particularly for an aircraft develop- ment program. This is true for most new experimental efforts, but is particularly true in V/STOL wind-tunnel testing, where the amount of available testing time can be severely restricted because of high wind-tunnel operational costs and costs required to support on-site testing.

The planning sequence will bediscussed brieflyasfollows:

1. Objectives of all testing

2. Testparameters to beconsidered

3. Model design

4. Wing-tunnel installation

5. Instrumentation and data acquisition

6. Program management

2.1 Objectives Definition of objectives for V/STOL wind-tunnel testing must include considerations in two areas: (1) establish the overall aerodynamic factors to be obtained from the test that will be essential to the pro- ram involved, and (2) obtain an overall feeling for and commitment of the support available for the program financial and otherwise). For a development program, for example, this can mean obtaining a decision based on studies as to how significant additional aerodynamic lift is related to the aircraft mission weight and cost.

Obviously the two are interrelated to the extent that in the end analysis, the available support establishes the complexity of the model and the type of wind tunnel that can be used. Examples of past wind-tunnel programs with various levels of each factor are shown below.

Factor Project support Aerodynamic scope Jet-in cross flow Component tests Minimal budget USB large-scale tests Complete model Low budget Tilt nacelle tests Component tests Low budget Grumman 698 Complete Large budget AV-8B 40 by 80 tests Complete aircraft configuration Large budget It should be understood that any of these cases could vary widely in model complexity and instrumentation requirements. The second factor, project support, will not be included in any further discussion but the principals in "aerodynamic scope" will be addressed. An additional item concerning the above test installa- tions is the possible justification for large-scale testing. This will be addressed in a following section.

2.2 Test Parameters For most wind-tunnel test planning, Margason (Ref. 5) has provided an excellent checklist defining the objectives of an investigation. As is usually the case for any wind-tunnel investigation of complete config- urations, the test results will include aerodynamic forces and moments for performance and stability. These characteristics may include aerodynamic propulsion-induced effects for transition flight (Fig. 1) as well as hover. Throughout this flight regime the test parameters must be chosen to define the complete flight enve- lope with assurance that the thrust effects are obtained. The configuration variables should be chosen and the construction of the model planned so that significant variables, such as flap settings or nozzle deflec- tions, can be evaluated. It is at this point that a judgment in the importance of exact duplication of all aircraft details be made. In planning for either a small wind-tunnel model or in large-scale testing (Ref. 4), this cannot be completely accomplished for a number of reasons; as a result, a concentrated effort should now be made to establish areas in which direct duplication of the full-scale aircraft are not essential for evaluating aircraft flight characteristics.

2.3 Choice of model It is at this point that the model design factors, such as model size, facility availability and model systems, should be considered as part of the planning. This will be directly influenced by factors discussed in appropriate sections of this lecture. However, there are several general considerations that can be made at this point.

2.3.1 Model size Anticipated scale effects can have varying significance on the choice of model size depending on how close to production the aircraft to be investigated is, how well defined the details of the final configura- tion are, and how significant the local areas of flow separation are in their effect on the data. For pur- poses of discussion, model scale will be referred to as small, less than 0.2 scale, moderate, 0.2-0.5, and large, larger than 0.5 scale. For exploratory investigations with simplified models and for low flap settings and angles of attack, and where accurate corrections are possible, small-scale models can be most economical and produce reliable wind-tunnel test results. This certainly was the case in the early AV-8B development (Ref. 8) in which a powered O.15-scale model was developed by modifying an available flow-through nacelle model. The resulting comparison with large-scale tests (Fig. 2) showed good comparisons with full-scale data in the linear angle-of-attack range. As might be expected, control and flap lift were slightly lower for the small-scale (Ref. 9) tests. More factors about both the large- and small-scale investigations will be dis- cussed later since they are classic examples of coordinating both large- and small-scale aerodynamic testing in V/STOL aircraft development.

As noted in Ref. 4, there have been several well-documented demonstrations of the significance of scale effects. These are reprinted in Figs. 3 through 7. For components of models, the actual size or local Reynolds numbers may be very small compared with that of the complete aircraft configuration. This was the case for the examples shown in Figs. 3-5, where the small-scale data indicated lower stalling angles for the inlet, deflector vanes, or leading-edge slats. On the other hand, there may be a "size effect" for the com- plete aircraft configuration such as the one for the small- and large-scale installations of Fig. 6 (deflected slipstream transport model). The large-scale model was powered by PW JT15D engines and tested in the Ames 40- by 80-Foot Wind Tunnel. The small-scale model was powered by ejector propulsion simulators, and the rela- tive size and shape of the 40 x 80 test section were simulated. The differences in the two sets of longitudinal characteristics, shown in Fig. 7, are probably not only scale effects but may be due to the nonsimilarity in both inlet and exit flow, particularly where jet efflux and its velocity profile effects the induced lift from the flap.

2.3.2 Power Systems The previous example of the importance of simulating the jet efflux in any powered-lift test is one of many which can probably not be completely documented but in which wind-tunnel testing produced erroneous per- formance estimations of the aircraft flight characteristics. As is discussed in Sec. 5.2, the current trend is to attempt to simulate the actual engine flow including bypass ratio. The CMAPS (see Sec. 5.2.4) is such a device on which its full application potential is currently being evaluated. The ejectors, developed in the 1960s and which powered the smaller model of Fig. 6 produced a uniform exit flow but were limited in exit pressure ratio and did not simulate the inlet flow rate for a given thrust level. A very valuable workhorse has been the tip-driven fans which have been built in several sizes up to more than 12 in. The driving factor in the use of any of these propulsion simulators is the need to get high-pressure air into the model and, if the forces and movements of the entire lifting system are required, that air must be "jumped" across the sting or wind-tunnel balance. The flexible lines, coil, or other transfer means can have significant effects on the design and cost of the model. At this point, if large- or full-scale testing is contemplated, several schemes of building a large but subscale model such as that of Fig. 8 should be considered. In this case, a model was built using 2-J97 turbojet engines which were 0.7 scale of a two-engine fighter design (see Sec. 5.2). Plan- ning for the power system must be integrated with other considerations, including the compatibility of the facility with the wind tunnel.

For rotary-wing projects, the power systems for some generic rotor tests are variable frequency electric motors. For example, the Ames 40 x 80 rotor test bed shown installed in Fig. 9 is driven by two 1,500 hp variable-frequency motors coupled together. Air-driven and hydraulic-powered motors are also used for small- scale rotor tests.

2.4 Wind-Tunnel Installation At this point in the planning sequence, the investigator must have a working plan that includes the aero- dynamic experiments needed to meet his objectives and must consider what model support and wind-tunnel test facilities are required. A table might be assembled similar to the one shown in Table I for in-flight jet noise simulation; it is described in Ref. I0. A section in this lecture covers typical wind-tunnel facilities that are available. Let it be stressed that for the particular demands of V/STOL testing, any choice should include consideration of the equipment and staff that could be made available to the program. Reference 11 or its equivalent will provide important parameters of candidate wind tunnels, but a meeting with the staff of each facility should be held to determine availability. Topics pertinent to V/STOL testing, among many other details should include: 1. Availability of electrical power or high pressure air.

2. Model motors and engine simulators available and maintenance required.

3. Capacity and currency of data-acquisition systems; the value of planned testing depends on reliable documentation of power for each test condition, and this is probably one of the most important topics in the _lanning.

4. Model support and balance hardware: for rotary-wing testing, few wind tunnels are currently equipped with all the support hardware needed- meaning added cost to the experimenter. For air-driven tests, rotary wing or otherwise, metering and air-transfer equipment are essential topics.

5. For acoustics tests, factors listed in Ref. 10 should be discussed.

6. For large- or full-scale testing, using gas turbine engines, specific items such as maintenance required, fuel, fire control, hydraulics, and start-stop requirements should be discussed. For the Ames 40 by 80, length of testing before temperature limits are reached is an important consideration, although planned improvements in ventilation will extend these limits.

After power systems and wind tunnels have been adequately considered, it is recommended that the combina- tion of model scale or size and type of wind-tunnel support be readdressed. The important factor of magnitude of wall-constraint effects for all projected test conditions may never be answered until after the testing is complete but, as noted in Sec. 5.1., advancements in computational fluid dynamics (CFD) are taking an increas- ingly essential role in this area. Information that could serve as a guide has been accumulated using the Ames 40 by 80 for tests on widely varying V/STOL aircraft configurations. The resulting wind-tunnel and flight-test data correlations were used to obtain the charts shown in Fig. i0 (Ref. 4 with G698 data added).

Data for both lift and momentum, as well as wing-span sizing, were obtained. The lines, representing possible size limits, have been somewhat arbitrarily drawn in the charts to represent guidelines for good wind-tunnel flight comparison under balanced conditions, and all data were obtained on complete lifting systems.

The possible acceptability of using partially complete models or component tests should be considered.

Most wind tunnels now in operation have a full stock of the hardware needed for semispan mounts and, if not, it usually is relatively inexpensive to manufacture the needed parts. Where lateral and directional character- istics are not required, the semispan mount is an economical method of studying high-lift characteristics, power-component performances, and loading on all parts. There is enough well-documented testing experience to ensure quantitative measurements and to evaluate most questions concerning effective aspect ratio. As examples, the large- and small-scale installations shown in Figs. II and 12 were tested in the Ames 40 by 80 and 7 by i0 wind tunnels, respectively. Objectives of both tests were to evaluate the performance and span- wise effects both in hover and during transition (Refs. 12 and 13). For the 40 by 80 tests, even though the effective wing-span-to-tunnel-width ratio was about 0.4 the wind-tunnel wall effects were minimal caused by the effective depth (80 ft) of the tunnel and the geometric aspect ratio, assuming the end plate (which was fixed and nonmetric) was the reflection plane. For the 7 by I0 tests (Ref. 13), the model size relative to the wind-tunnel cross section was much larger, but quantitative measurements were still obtained on the ejector performance and its influence on the external aerodynamics of the wing. For this case, accurate evaluation of hovering performance was obtained in the wind tunnel with the large side access door open and blowing the ejector exhaust into the shop area.

2.5 Instrumentation and Data Acquisition During the foregoing considerations it must be assumed that model instrumentation and data acquiring and processing have been assigned significant roles in the planning. The continuing development of microprocessors and more economical computer systems must be considered in choosing the data-acquisition objective. For example, it is becoming mere feasible to obtain a significant amount of pressure data along with the overall force and moment data and a few directly measured component loads. Nevertheless, as indicated by the possible measurement needs for a V/STOL investigation listed in Fig. 13, the advantage of using large-scale solely for the purpose of storing on-board data-acquisition equipment may still offset possible higher model construction costs for the large models. To obtain all the items that can be evaluated on board one large-scale model, it is possible that two-small scale models would have to be built or that at least a force and a pressure model would be built. It is, therefore, essential in planning to evaluate the state of technology of such items as electronic Scanivalves, straingage or capacitance balances or both, or flow-survey equipment.

Most established wind-tunnel facilities are continually evaluating and updating their data acquisition equipment and, during discussions with the wind-tunnel staff the potential limits on instrumentation should be thoroughly discussed. Since current V/STOL testing taxes the most up-to-date computer system, it has been my experience that a program can be enhanced by the experimenter using an auxiliary small computer for some of the measurements that can be correlated using a time-code with the data reduction supplied by the wind tunnel. For rotary-wing testing, dynamic analysis hardware and software are now available at most of the larger wind tunnels. At the NASA 40 by 80, this system is on-line and is an essential part for a safe and successful test procedure.

2.6 Test Program Having obtained the above information and made the decisions and commitments, a written test plan should be produced and distributed to interested parties to ensure that agreements were as perceived. Except for smaller programs, which are funded and activated in-house, this is good practice for any experimental effort, but it is essential for V/STOL testing to the extent that all those designated in Fig. 14 be informed. The particular examples shown in Fig. 14 were large-scale test programs, but they are still representative in terms of the significance of planning and coordinating most V/STOL wind-tunnel test programs. They are presented in the order of increasing program costs with the numbers listed corrected for 1983 cost-of-living levels with respect to the level existing at the original test dates. The following is a brief description of each as reprinted from Ref. 4.

The test results for the project of Fig. 14(a) are reported in Refs. 14 and 15. This was an internally managed, basic-research, project with the objective of studying stability and control, high-angle-of-attack characteristics through stall, and acoustics. Project management came from Ames Low Speed Aircraft Research Branch (FNA) with both an engine and an airframe manufacturer acting as consultants. Final reporting was done by Ames.

Test results for this project shown in Fig. 14(b) are reported in Refs. 16-18. This was a U.S. Navy-NASA funded program, managed by NASA, with the objective of obtaining static (in and out of ground effect) and wind-tunnel data for loads, stability and control, and performance. The model was heavily instrumented to document propulsion and external-surface pressures. The test management came from the Ames Aircraft Project Office, with the FHA serving as advisors. The airframe contractor supported the tests with design, test support, data analysis, and reporting.

Test results for the project of Fig. 14(c) are reported in Ref. 19. This was U.S. Navy funded with test management coming jointly from the FHA, the contractor, and the Navy. The model was equipped primarily for full-scale stability, control, and performance checks with the instrumentation required to document power settings together with a few loads. The cost of the model was an order of magnitude higher than those in the foregoing projects because the model combined a fuselage flight structure with the Rolls Royce F402-R-402 engine and a boiler plate wing-flap system. In addition, since the tests were part of an aircraft development program on a tight schedule, it was highly "visible," and a large number of contractor personnel were required to support the operation of the tests and correlate test results on a daily basis.

Each of these projects has proved to be productive, with the first program being the forerunner of the NASA Quiet Shorthaul Research Aircraft (QSP_A); the last program is currently near the production stage. All projects included considerable wind-off testing, as well as wind-tunnel work, and measurements obtained were in the following areas of study.

1. Stability and control 2. Aerodynamic performance 3. Propulsion performance, inlet and nozzle 4. Loads on flaps and control surfaces 5. Surface-pressure measurements on all components 6. Boundary-layer surveys 7. Acoustic studies, near- and far-field 8. Wake surveys, downwash, sidewash 9. Static pressures on tunnel walls to evaluate wall effects 10. Flow visualization 11. Structural static and dynamic loading 12. Aircraft systems check All of these topics plus those pertinent to rotating-wing systems continue to be subjects of V/STOL wind- tunnel testing, and the remainder of this lecture will address details to be considered in their measurement.

3. WIND TUNNELS This section presents a short survey of available wind-tunnel test facilities that have, in general, large test sections, are subsonic, atmospheric, and have sufficient support equipment and personnel for test- ing a complete, powered V/STOL aircraft model through transition speeds. The section is not intended to be a complete survey but rather one that indicates types of wind tunnels and equipment that are available. For this reason, only one large industrial tunnel is mentioned; only the high-speed and pressurized wind tunnels at Ames are listed, though they are not described in detail.

For evaluating the capability of a wind tunnel for V/STOL testing in which flow separation may be occurring at anytime during normal flight operation, it could be assumed that the flow quality in the test section may have a large effect on the power-induced aerodynamics. Because of this, at most of the large wind tunnels, considerable effort has been made to evaluate and improve these characteristics. The significance of flow qualities is discussed in Refs. 20-22 but, for V/STOL, they should include turbulence as well as uniformity of test-section dynamic pressure. Unfortunately, except for the DNW and Langley 4- by 7-m tunnels, complete documentation of these characteristics was not available at this time.

3.1 DNW Wind Tunnel The DNW facility is probably one of the most recently developed. A general description appears in Ref. 23, and a compilation of calibration data is included in Ref. 24. As indicated in Fig. 15, the wind tunnel is a closed-return design and has interchangeable test sections. The latter include 6 x 6 m, 8 x 6 m, and 9.5 x 9.5 m sections as well as an open section for acoustic testing for which longitudinal slots are provided in the walls permitting an open-to-close area ratio variation up to 12%. Flow qualities measured to date are excellent with a deviation of static and dynamic pressure across the test section of less than 0.3% flow angularity ±0. I °, and a turbulence level of 0.1%.

The facility has hardware for either floor or sting mounting, l_qe sting has sufficient degrees of freedom to allow strictly vertical movement without pitch change or pitch and yaw without vertical or lateral displace- ment. The hydraulic actuation allows a downward velocity of 5 m/sec at 0.5 g for possible use in flare simu- lation. As with most modern sting arrangements, a high-pressure air-supply line is included to supply com- pressed air to the model. Floor mounting to the external balance is p_ssible for either strut-supported or semispan models. For gound-effect tests, a 6.3 by 7.6 m (width to length) moving belt can be installed in the two large test sections with a belt speed of 5-60 m/sec. The open-throat feature has been designed for use in acoustics studies. An acoustic evaluation of the facility is reported in Ref. 25.

3.2 8-m ONERA Transonic Wind Tunnel The ONERA facility has been operating since the 1950s in Modane, France (Ref. 26). The Mach number range is from 0.03 to 1.02, which might enhance its utility for V/STOL testing by making it possible to test with the same model installation from low transition airspeeds to transonic speeds. Although, as discussed in Sec. 5.3, it may not be practical or economical to do this. The facility has a round test section but is equipped with a blown ground board for handling strut- as well as sting-mounted models. An important feature affecting test operations is the ability to exchange test sections, with options such as those shown in Fig. 16(b). For complicated powered models with a network of instrumentation needed, this makes it theoreti- cally possible to make all the pretest checks while another project is occupying the tunnel.

3.3 Langley V/STOL, 4- by 7-m Wind Tunnel The circuit of the Langley V/STOL facility is shown in Fig. 17. The test section is 4.42 m (14.5 ft) high by 6.63 m (21.75 ft) wide and 15.2 m (50 ft) long. The maximum speed is 103 m/sec (200 knots) and it can be operated in a variety of configurations - closed, slotted, partially open, and open, the latter being open on three sides. Studies have been conducted on methods of improving the flow quality in the test section (Ref. 27). Nevertheless, the wind tunnel has continued to be an essential tool in NASA's low-speed experimen- tal aerodynamics studies and has most of the hardware required for supporting models either by a sting or by using a floor-mounted strut. A moving-belt ground plane is also available. The effort to improve the flow quality and productivity will take a significant step when the wind tunnel undergoes major modifications in late 1984.

3.4 Boeing V/STOL Wind Tunnel The Boeing tunnel (Fig. 18) is a typical industry-managed facilitywhich is kept uo-to-date as required for company developmental testing, as well as for industry-wide and government-supported programs. There is an option of a 6.6- by 6.6-m (20- by 20-ft) or an open-throat test section, as shown in Fig. 19, with the test section open on both sides. The particular model installation shown is for a small-scale aero-acoustic test for which the model design is discussed in Sec. 4.1.3. This wind tunnel also has the options for either strut- or sting-supported tests. At present, the tunnel is capable of maximum speeds of 250 and i00 knots for the closed and open test sections, respectively.

3.5 Ames 40- by 80:ft/80- by 120-ft System The Ames 40- by 80-ft has been in operation for about 40 yr, but a recent project (described in Ref. 29) has extended the speed range and added a new test section installed in another "leg" as shown in Fig. 20. To do this, a new drive system has been installed, and vanes are being designed to divert the 80 by 120 test section inlet flow into the fans through the new inlet. It is too early to detail all features of operating in the 80 by 120 test section but the 40 by 80, as it was originally conceived, has proved an invaluable tool for investigating V/STOL aerodynamics, as well as flight-like aircraft systems. With reference to the 40 by 80 test section shown in the elevation view (Ref. 30) of Fig. 21 (updated to show recent changes), the large (and sometimes very heavy) models or aircraft are moved into the outer "high bay" ground level or "2 ft level" and, when readied, hoisted up over the clamshell doors onto the support system. A three-support model instal- lation (the full-scale model of the Grumman tilt-nacelle configuration G698) is shown in Fig. 22. A sketch of the 80 by 120 test section is shown in Fig. 23, viewed from outside the test section.

The sizeof such a testsection asthe40by80has allowed special studies and testsonaircraftcom- ponents; for example, theacoustic investigation withaninstallation shown in Fig.24(a) (see Sec. 6.2.2) and thesmall-scale hover test in Fig.24(b).In thelatter case a "flying"ground plane was used to incorporate pitchand roll effectsintothebasic ground-effect study.The data-acquisition support hardware and power systems have been gradually updated through theyears, and current data systems include anon-board system thatcan bemade compatible withflight-testinstallations.Most large- and full-scale models can onlybe practically supported from below, butformoderately sized models, sting-support hardware has recently been made available.Since aircraftengines at this scale arecommonly runin thetestsection, available support systems include those for fuel, C02 and engine servicing equipment, and thepower systems usually available, such asair supply and variable frequency power. The facility has a group of qualified aircraftmechanics.

3.6 Langley 30-by60-ftWind Tunnel

The Langley 30by60is a classic large-scale installation, and has been in operation since the1930s.

The circuit is shown in Fig.25. Asshown in Fig.26,it is anopen-throat tunnel withthemodel located just in frontof thefans. It has been used for thedynamic stabilityand control studies described in Refs. 31 and 32and discussed in Sec. 4.4.1. The wind tunnel has been utilizedquiteregularly in aviation development aswas thecase for theinstallation in Fig.26. Unfortunately, because of theporximity of thetestsection to thedrivefans,validacoustic testing has been difficult.

3.7 NRC 9-m V/STOL Wind Tunnel

Except for DNW, theNRC 9-m wind tunnel (Ref. 33)is one of thenewest low-speed tunnels, having been in service since 1970.Its development included testing of a i/lO-scale model of thewind tunnel and, although quantitative values of flowturbulence arenotavailable for publication, thereis a setof screens immediately upstream of thetestsection.Most of theV/STOL testing experience to date hasincorporated a floor-mounted strutsystem such astheinstallation shown in Fig.27,which is connected to thebalance system below the floor. One particular asset for air-driven V/STOL testing is thecow,pressed air supply.It has nearly full use of theNRC Blowdown Wind Tunnel supply which produces dryfilteredair at 2050 kPa (300 Ib/in.2). Atthe wind tunnel, this means 1,700 kPa (250 Ib/in.2) for 30kg/sec (45Ib/sec).Unfortunately, heating capability is limited.

3.8 Other Wind-Tunnel Test Facilities

The foregoing briefsurvey has identified some of thelarger wind tunnels thatareatmospheric and have complete facilities for both fixedand rotary-wing investigations. Not described were theUnited Kingdom large subsonic tunnels, theLockheed tunnel, thehigh-speed tunnels such astheAmes Unitary System (Fig.28), and the12-ftpressure tunnel.The ll-ft testsection of theUnitary System is capable of holding a continu- ousspeed down to M=0.4 sothata program can cover subsonic airspeeds for models designed for transonic and supersonic testing. The CMAPS program, at Ames (see Sec. 5.2.4)is designed to make such testing evantu- ally practical byaccurately simulating inlet and exit flows. Inaddition, therareseveral 7 by10wind tunnels in academia, government, and industry which can beuseful for small-scale or component testing.

One of themost economical systems of wind-tunnel testing, particularly asit is applied to V/STOL test- ing,is theone described byKnott (Ref. 34)and operated byBritishAircraftCorporation (BAC), Military AircraftDivision (Fig.19). The testsection is 5.5m(18ft) wide and 5.0m(16.5 ft) high, butthesystem operates over a narrow speed range of II to 21m/sec (35to 70ft/sec). This requires aninstalled power of only200 kW (250 hp). There is some disadvantage in thattheexisting installation is theopen-throat, no-return type and its test-section flowqualities are,therefore, subject to weather. The design trade-off may, then, beone of flowquality required against thecostof screens and flowstraighteners resulting in increased power required. The testing and scaling techniques assume thatMach number simulation is not important. Table 2 shows typicalvalues for thetestvelocities compared withfull-scale values.Amajor advantage could belower model costs and lower on-board power requirements. This alsocould mean a "lower profile"support system required to take thelower model loading.

4. TESTING INSTALLATIONS FOR V/STOL

Since thechoice of theproper testinginstallation depends onmany factors - nottheleastof these is cost- it would bemeaningless to suggest any given procedure for choosing themodel installations. The following areexamples of testprograms thatin most cases, proceed through a variety of wind-tunnel test setups.

Much information can beobtained from such exploratory testsasthatshown in Filg. 30(Ref. 35),which had theobjective of evaluating characteristics in hover and forward flight for a ground-effect vehicle-the "Avro" car. This test, in theAmes 7- bylO-fttunnel was made after40by80testsshowed thefull-scale vehicle had a large aerodynamic center location change asit lifted outof ground effect. The small-scale testwas a very low-cost effortto investigate thefundamentals. Asampling of thetestresults is shown in Fig.31. Ground effectwas measured very close to theground board.Although notdone, thedata could have been corrected for thedisplacement thickness of theboundary layer bythemethod similar to thatsuggested laterbyEast (Ref. 36): _* aC a_* aC Ccorr = Cmeas x TT_+-_X- _- _ where the gradients of C are determined by carrying out additional tests with other boundary-layer thick- nesses. No moving belt was needed with BLC being planned for but not used and qualitative correlations were made with the full-scale data.

An example of a more sophisticated (and more costly) test is that of the Grumman 698 tilt-nacelle devel- opment. It was decided that a full-scale test was needed to verify control which might be available from the 698 control-vane effectiveness design. A component test installation in the Ames 40 by 80 tunnel was used (Ref. 37). The nacelle was powered bytheQ-fan (Ref. 38),and was mounted onthe40by80wind-tunnel turn- tableasshown in Fig.32,which provided thenacelle witha large angle-of-attack range.Inthis case the inlet and fanitself were thesubject of another investigation (Ref. 39),sothattwo testobjectives were met using common hardware.

Athirdexample, thatof thedevelopment of theXFV15 tilt-rotor aircraft(Refs. 40and 41),included the series of testinstallations shown in Fig.33. Going to a higher diskloading and amore usable drivesystem meant a series of testsvarying from small- and large-scale semispan models to testing theactual aircraftin theAmes 40by80tunnel.

The range in possible types of installations varying in complexity and costbetween those of theprevious examples is large,and thetestobjectives arediverse.Inthefollowing discussion, experience withtest installations varying in type and complexity will bediscussed in more detail.

4.1 PartialInstallations

Even though for V/STOL investigations it is usually essential to simulate thecomplete aircraftconfigura- tionin thefinal aerodynamic evaluation of theaircraft,component testing, two-dimensional testing, and quasi-two-dimensional testingcan take a major partin theprogram. Examples of these installations follow.

4.1.1 Component testing

Inone of theprevious examples (Fig.32)it was necessary to establish thecontrol effectiveness of the pitch-control vane before going ahead witha full-scale model. In thattest, thedata of Fig.34verified predictions thatthefaneffluxwould provide sufficient control effectiveness through a wide range of nacelle attitudes.The vane loads and fanflows were documented withsufficient on-board information to establish design information for proceeding withtheprogram. Note thatin this case, it didnotseem necessary to simulate thewing or fuselage interference effects. This would bethesubject of thenext phase- a full-scale model anorder of magnitude more.costly than this one.

For a two-engine fighterconfiguration, overhead inlets have recently been investigated bySmeltzer et al.

(Ref. 42)and Durston and Smeltzer (Ref. 43),using a component or partiallycomplete model installation, as shown in Fig.35. The inlet region of themodel was sufficiently instrumented to obtain both external flow- field contours forward of theinlet and inlet distortion and pressure recovery. This was done using five-hole cone probes. The forward partofthemodel was sufficiently simulated to relatechanges in leading-edge con- figuration to flowat theinlet plane for both subsonic and supersonic speeds. The model had flowthrough the nacelles, butmetering plugs were added to vary flowthrough theinlet.

Certain basic Study projects and key experiments can alsobeplaced in this category. Many researchers have been accumulating wind-tunnel data onthejet-in-cross flowproblem, asreported in Refs. 44-46 for the installation of Fig.36. Initially, this type of experiment was started withtheuse of a flat plate,as shown in Fig.37,witha 10-cm (4-in,)jet issuing from thecenter.Then, using a smaller testsection (Ames 7 by10),thebody of revolution and another flat platewere tested.All testshave been closely coordinated in objective and scope, thatis, to evaluate thepressure onthesurface surrounding thejet exit and to study theflowin thejet itself. From this data base, several jet models of thejet-in-cross flowproblem arebeing evaluated, among them being thediffuse vortex model of Fearn and Weston (Ref. 45). The 5-cm (2-in.)jet model has alsobeen used recently to study theflowcharacteristics using a three-component LDV technique which is described in Sec. 6.3.2.

4.1.2 Two and Quasi-Two-Dimensional Testing

Before thesignificance of powered-lift effects onairfoil section characteristics was fully realized, airfoil sectional characteristics were accurately evaluated using test installations of constant chord and spanning thewidth of thewind-tunnel testsection.The height of these tunnels was generally much more than theirwidth to minimize blockage corrections. For high or powered lift airfoils thetechnique has been suc- cessfully applied using precautions for preventing flowseparation ontheend wall. Inone case (Ref. 49)for theinstallation of Fig.38,a powered nacelle was added to a two-dimensional high-liftexternally blown flap (EBF) installation to evaluate leading-edge stall control and turning effectiveness. Asin thecase ofany airfoil adjoining a flat plate,theadverse pressure gradient ontheairfoil is transmitted to thewind tunnel or end-plate walls. Apositive way of controlling this is to uselocalBLC at critical locations on theplateitself.

This was alsodone to thequasi-two-dimensional model shown in Fig.39. Inthis instance, theobjective was to study effects of airspeed ona span-wise ejector (Ref. 50)and to derive thetwo-dimensional section characteristics of such a high-liftdevice for correlation withlarge-scale testsona complete model config- uration.It was found thatalthough thetwo-dimensional effective angle of attack could notbeevaluated to much closer than ±2 °, theeffective aspect ratiowas sufficiently large, - approximately 8 (based onthe power-off drag polar) - to make thepower-on data useful in evaluating theejector performance asfunctions of bothpower and angle of attack.The key to such aninstallation is thedesign of theend plate. There are very few published data onlarge end plates except those byRiley(Ref. 51);more recent studies arebeing directed at drag reduction of higher aspect ratiowings.Fortheinstallation of thehigh-speed evaluation of

wing cruise blowing system byMahal (Fig.40;Ref.52)theangle of attack was evaluated as

=_geom - k Cn

k =3.54 - 1.26(1 - M2) I/2

Thisis consistent withall thepower-on data obtained, and was used in evaluating theeffects of blowing on thedrag riseMach number at a targetcruise lift. Thisinstallation shown installed in theBoeing wind tunnel (Fig.40)had anend platethatwas contoured to minimize drag at high subsonic speeds. All data were in the form of surface pressures integrated to givelift and pitching movement withmomentum rake data being used for drag evaluation.

4.1.3 Semispan Installations Semispan models have been widely used in the conceptual stage of a high-lift concept. In the 1950s very economical installations, such as the one described by Anscombe and Williams (Ref. 53) and shown in Fig. 41 were developed. The advantages over complete aircraft simulations are lower cost plus larger scale potential for a given test-section size. The main disadvantage is lack of any lateral-directional information and, possibly, questionable pitch control-data. Most installations rely on the use of either a mechanical or electrical balance, located below the floor.

A major choice for such an installation is whether to have the model include a half fuselage that is metric, such as the one in the sketch of Fig. 41; a.fuselage that is nonmetric, or attached to the wind tunnel floor; or an end plate that is nonmetric with the wing protruding through it. Except for exploratory tests in which the longitudinal characteristics of a canard/tail-wing configuration are needed and each component must be repositioned, the author is partial to the use of the metric half fuselage. Experience in the Ames 40 by 80 with the semispan installation of Fig. 11 (Ref. 12) has been excellent to the extent that approximate correla- tions were made with a complete jet flap wing-fuselage model. For the installation of Fig. Ii, the root or effective plane of symmetry was raised to be clear of the floor boundary layers. The test reported in Ref. 54 and shown in Fig. 42 incorporated a half fuselage that pitched with the wing but was nonmetric and attached to the floor fairing of the tunnel. The disadvantage of this is sealing problems where the root of the wing had to be nonrestrictive in order to avoid affecting the measurements of the wing characteristics. For high-lift conditions, the pressure difference places an additional force on any seal that is installed.

The semispan installation was also used economically in the development of the XV-15 tilt-rotor model mentioned earlier. Using the installation in Fig. 33(c), both power-off and power-on tests were made as described in Ref. 55 to evaluate rotor aerodynamic performance, rotor dynamics, and hardware of the rotor sys- tems. Even though the root of the wing was rigidly fixed, the wing-rotor dynamic measurements could be used to predict operating limits of the complete aircraft. The model shown was unpowered and could be assembled with spars of three different natural frequencies. During the windmilling tests, the effect of damping of the rotor-pylon-wing twist mode was evaluated with step inputs into the rotor tilt actuation system. The rotor performance was later studied using a dynamometer test installation. All test results were correlated with small-scale measurements and predictions.

4.2 Testing Complete Aircraft Configurations Since the wind-tunnel time for complete aircraft configurations is a major part of the aerodynamic and dynamic development for any aircraft, conventional or V/STOL, any cost saving in this phase can have a major effect on the overall development costs. Powered V/STOL models themselves are inherently more expensive than unpowered conventially configured models. This is a result of additional costs that are incurred not only because of adding power to the model but also because of the additional effort needed to support the tests.

Other costs come from the increased amount of on-board instrumentation for which leads have to be "funneled through" or around model supports. As for moving-base or for free-flight operation, the model must be suffi- ciently light and designed for adequate remote control. Low model weight and light power units are also required for track-supported models such as used at the Princeton track facility. Examples and problems of each of the above will now be discussed.

4.2.1 Sting Supports For the rigid mounting category, sting supports, in addition to high-speed requirements, are sometimes desirable for minimizing aerodynamic interference induced by model lift, particularly in ground effect. For some test conditions, strut mounts can induce buoyancy or blockage under the model that may be difficult or impossible to evaluate. For V/STOL configurations, it is important that the sting support be downstream of the area generally occupied by lifting jets and flaps. As can be seen by the model support in Figs. 43 and 44 (Refs. 56 and 43), however, the rear end of the model can be adapted to receive a sting of sufficient size to minimize support flexibility and, at the same time, enter the model with a minimum alteration of the rear contour.

In the case of the CMAPS-powered model shown in Fig. 45 (Ref. 57) for the two-engine supersonic fighter, the pressure and instrumentation leads were routed forward under the fuselage. During these tests an attempt was made to set the narrow segment of sting as short as possible to reduce "bounce" and length of high-pressure leads. Lower surface pressures were calculated at the model plane-of-symmetry; a summary of the results is presented in Fig. 46 (Ref. 58). It is evident that lengthening the sting or moving the adapter 0.30 m (12 in.)

farther from the model reduced the influence of the adapter to a very small amount (_Cp = 0.05) at the model trailing edge.

For some powered fighter configurations, an off-side or overhead mount could minimize restrictions in direct simulation of the rear of the aircraft such as the mount shown in Fig. 47 for a generic fighter investi- gation. Instrumentation leads and the air for powering the propulsion jets were routed in through a twin support at the vertical tails, thus leaving the lower-rear of the model free of any induced flow generated by the deflector nozzles. This was mandatory in this case because the objective of the test was to evaluate the effect of nozzle configuration on the aerodynamics of the model over a large Mach number range.

A somewhat larger model built for lower speed with a corresponding lighter construction (lower weight per size) was tested in the Ames 40 by 80 as shown in Fig. 48 (Refs. 60 and 61). This was a sting-supported model to be installed on the 40 by 80 turntable, and the model chord plane was vertical so that rotation of the turntable would change angle of attack through high angles. The configuration was a side-by-side twin-engine configuration so that the sting installation required a widening of the rear fuselage slightly beyond scaled width. For this case, it was judged acceptable since aerodynamic corrections could be made using integrated surface pressure measurements. Also, in this case, it is believed that the thrust of the fans had little interaction with the sting since they were used to pump air through the inlets, and the momentum of the exhaust was exceptionally low. A disadvantage of this installation was the asymmetrical lateral/directional effects on the model induced by the vertical strut.

Asthefull model sizecapabilities of thelarge atmospheric wind tunnels areutilized,themodels may become large and heavy enough to preclude stingsupports. Atthis stage available stinghardware may produce mounting thatis tooflexibletomaintain a stable platform for thebalance. There is probably some model size-weight levelfor which strutmounts become mandatory for practical reasons and for which therequired stingsizebecomes toolarge to allow tailoringtherearof themodel to scaled aircraftcontours. There appears to benocoherent criteriato help in thestrutor stingchoice buta thorough study of theinteraction of possible oscillating aerodynamic loads and themodel-support dynamics is anessential partof this choice.

4.2.2 Strut-Supported Installations

Fora configuration in which there is a large amount of power-induced circulation from thewing, the single support mount shown in Fig.49(a)(Ref. 62),a i/7-scale testof theH126 jet wing aircraft,was used successfully. Forcomparison purposes, theaircraftitself is shown in Fig.49(b) mounted in theAmes 40by80.

The small-scale testwas made in theAmes (Army) 7 by10and employed a pitchlink activated bya rodrunning parallel to themain strut. The large fairingenclosed instrumentation and thepitch-link actuator.Inmost cases, this actuator can belocated below thefloorto reduce therequired fairingsize. When thelift jet effluxis in thecenter of themodel or it is thought thatstrutinterference effects would beexcessive, the model canbeinverted, asshown in Fig.50(Ref.63). Aswithstingmounts, a disadvantage of theheavier models is thatthesingle-strut installation tends to bemore flexiblein theyaw-roll modes of oscillation.

An interesting typeof hybrid mounting system (Fig.51)has been used byBoeing Vertol for rotary-wing testing. It incorporates theair-driven "power pod" in supporting a helicopter model. The pod itself has been attached to thestingwhich is equipped for pitching and yawing theentiresystem. The "pedestal" or vertical strutsupports themodel through thebalance inside themodel. An optional setup uses thesame power pod for a floor-mounted installation.

Because of model weight, large- and full-scale models at thelarge-scale testfacilities (40by80and 80by120 at Ames and the30by60at Langley) must use floor-mounted_struts, usually, a three-strut system as shown in Figs. 22and 52,respectively. The strutstend to bestiff in yaw if themain strutsaremounted to theouter wing, such asin thecase for theAV-8B model of Fig.53. Acompromise was made in thelanding- gear-type mount for theSTOL t_ansport model shown in Fig.6(a)(aswellaswiththe698 model of Fig.22).

Forthese models there was a special effortto design themodel withthestrutsaway from theundersurface of thewing in order to minimize anylocaleffectonwing airflow separation.

Even withtheeffective thickening of thestruttips due to attachment of leads, there has been a wealth of evidence thatfor theV/STOL configurations tested to date,, strut-interference effects have been extremely small.Thisconfidence has come primarily in thecomparYsons of full-scale wind-tunnel testdata for actual aircraftwiththeir corresponding flight-testresults (Refs. 64and 65). Data representing anassortment of V/STOL aircrafthave correlated wellin both angle of attack for a given lift and power setting, aswellas for stabilityand control. Some of thedata for thesmaller aircraft,such astheVZ3 and XV3, correlated wellwithout applying any wallconstraints or blockage corrections to thedata. Concerted effortsto general- izewhat struttares exist,assuggested in Ref.6, have notbeen successful for V/STOL models, particularly at thelow-speed end of transition withhigh flap settings and jet deflections.

4.3 Ground-Effect Testing

Asshown bytheexperience withevaluation of theflight versus wind-tunnel data of theAugmentor Wing Research Aircraftat Ames and discussed byCook and Whittley (Ref.64)a highly deflected jet impinging onthe floorof thewind tunnel must beaccurately simulated. It must beprevented from migrating toorapidly forward through thelowmomentum of thefloorboundary layerthen blowing back up(which may or may notbethecase for theactual aircraftin flight) and overthewing.Thiscan bereasonably wellsimulated bymaintaining a small boundary layerbyeitheror both BLC and moving belt (Ref. 66). An additional method is themoving-model technique such ashas been used atthePrinceton and other towing facilities (Sec. 4.4.2). In1977, Campbell et al. (Ref. 67)reviewed thestatus of theuse of wind-tunnel measurements to predict aircraftflight charac- teristics,including initial experience withtheC8A just mentioned. Since thattime,similar comparisons have been made using both theHarrier and QSRA, results ofwhich have notbeen welldocumented. Most of these com- parisons varifytheso-called Turner criteriawhich is shown in Fig.54(taken from Refs. 29and 68)or the equivalent if BLC is used to control theboundary layerontheground plane.Margason (Ref. 5) has also reviewed theneed for a moving beltor BLC and concluded thatfor low-aspect-ratio wing configurations, partic- ularlythose using deflected thrustconcentrated in asmall area, lesscontrol of thewind-tunnel floor boundary layeris required.Inany case, some of theflowphenomena restricting themeasurement of ground effectof V/STOL configuration in a wind tunnel arethesame asthose governing model sizing (Sec. 5.1).

Turner's beltinstallation (Ref. 68)was in anenlarged inlet of theLangley 7 byi0 300-mph wind tunnel and, asshown in Fig.55,is close to theinlet of thetest section.The resultwas asexpected. Asshown in Fig.56,withthebeltspeed synchronous to thetunnel airspeed, theboundary layer was completely eliminated.

The earlydevelopment of themoving beltitself, asindicated byButler et al. (Ref.69),was tedius, but through advancements in materials and multiple drivesystems, current operation is relatively trouble free.

Atthepresent time,most of themoderately sized low-speed tunnels having stingsupports areequipped witha moving beltsimilar to theone in theBoeing Vertol tunnel (Fig.57)where theessential parts of any belt system arenoted.Asshown, thesuction slot provides BLC forward of thefrontroller.

Amore economical method of accounting for thewind-tunnel boundary-layer growth is theplacement of BLC onthefloorbyitself or dombined witha raised ground board forward of themodel. Astudy was completed for NASA in 1974 (Ref. 70)to explore thefeasibilityof a BLC installation in theAmes 40by80wind tunnel.The objective was to facilitateground-effect testing, aswellasto lower thetesting airspeed restriction for propulsive systems withhighly deflected jets. The finally recommended design was notused for economy reasons, butthedesign (Fig.58)is still considered valid,and basic aerodynamic studies byHackett et al., which were vital in understanding BLC requirements, continued through the19_0s. Other investigations were summarized in a 1973 lecture byPoisson-Quinton (Ref. 71). Inmost of thestudies, small-scale jet flap testswere made, using both a beltand BLC. Typical wind-tunnel results, though preliminary, were updated by Hackett et al. in Ref.72,and a sample setof pitching-moment data withtail onis shown in Fig.59for the case of 2 chord lengths above thefloor. The correlation between fixed-ground plus BLC and themoving beltare good, even though, at Cv =3, thejet wake is probably startingto impinge onthefloor. When this happens just behind anunswept wlng or two-dimensional jet flapmodel, theflowpattern, which includes a standing vortex, is established and causes suck down. The control and intensity of this flowpattern is affected by floorBLC (ormoving-belt velocity), butmay notgetestablished during thelanding flareor any other transient situation.

4.4 Moving-Base Testing

If theabove mentioned flowpatterns arenotparticularly Reynolds-number sensitive, moving-base tech- niques may beused to evaluate theireffects. V/STOL models configured to simulate thepower loading asa complete aircraftconfiguration and at thesame time to have sufficient scale to ensure quantitative test results tend to become heavy. An example of hardware needed to move a heavy model in vertical motion is shown in Fig.60,which was theresultof a design study (Ref. 73)to investigate problems of simulating thetran- sientground effects in theAmes 40by80. The design criteriafor a O. ll-scalemodel of a large STOL trans- portaircraftincluded thefollowing testconditions and model parameters:

Dynamic pressure, kPa (Ib/ft2) 1.93 (40.3)

2.24(10) 6

Reynolds number, RN

Model weight, KN (Ib) 3.23(727)

Model span, m(ft) 3.7(12.1)

Sinkrate,m/sec (ft/sec) 5.2(17)

Rotation rate(forflate) deg/sec

Scaling full-scale flareand touchdown maneuvers brought large accelerations; theresultis thedesign of Fig.60and prohibitive costs.Even though atrade-off between facility cost and model scale may reduce the sizeof thesupport somewhat, studies of some of thequestions ontransient ground effects and aerodynamic damping using scaled experiments may still have to relyontheuse of trackfacilities and free-flighttech- niques. The following is a briefdiscussion of twosuch facilities.

4.4.1 Free-Flight Methods

Although theterm free-flighttestingis alsoapplied to high-speed or ballistictesting, or thespin- tunnel testingfor V/STOL eva_uatilon, it has usually become synonymous withthewind-tunnel free-flight methods such asused in theLangley 30by60. With use of this technique, aqualitative evaluation of flight charac- teristicsof awide range of aircraftconcepts has been made in and outof ground effect. The well-known technique (Fig.61)incorporates Froude-scaled replicas of theaircraft. Reference 74includes a concise description of significant factors of thetechnique, and equations of motion for themodels arelistedin Ref.75. For most current aircraftdesigns, thegeometric scale tends to runfrom 1/10 to 1/6,withlengths 2.1m(7ft), wingspans of 1.52 to 1.83 m(5to 6 ft), and weights of 224 to267 N(50to 70Ib). With this scaling, themodel angular motions areupto 3 times asfast asthose at full scale.Thisbrings therequire- ment for multioperators. Since theoperators must attempt to keep theaircraftin one position or ona pre- determined flightpath and, asmentioned, thetime-constants are1/3full scale, light, quick response actuators have been developed which provide thepilot withtight control.The model is instrumented tomeasure linear and angular acceleration along withcontrol-surface positions which aretransmitted throughout theflight to strip charts.The flight cable alsosupplies thehigh-pressure air to themodel motors. Motion pictures document theflightpath and representations of desired automatic stabilityand control augmentation systems areinputintothecontrol systems.

Although these techniques have gradually been improved through theyears, Paulson (Ref. 76)still gives one of themost detailed descriptions of thetechnique and its problems for use withtheB-58.Ina more recent application, thefree-flighttechnique was used to evaluate thehigh-angle-of-attack characteristics of a forward-swept wing fighterconfiguration (Ref. 31). The program included theuse of two models, one a O.16-scale model of thecomplete aircraftand a O.16-scale flat plate model. The former was both fixed-support tested and tested in freeflight. The project was coupled withtestsof both models ona free-to-roll apparatus in order to evaluate any unclamped roll oscillations at high angles of attack.The model is shown in Fig.62being tested in thefree-flightmode. Damping in roll evaluated from thefree-flighttests was found to agree quantitatively withtheflat-plate measurements for moderate angles of attack.

4.4.2 Moving-Base Track

Amajor difference between thefree-flightfacilities and themoving-base trackis thelatter'scapability of restraining certain undersired degrees of freedom of themodel motion. Therefore, it pre-programs a flight- path such asa landing flareor a roll oscillation. Inthecase of thePrinceton Track (PDMT), Froude-scaled models aremoved along a pre-programmed route thatis 246 m(750 ft) long (Ref. 77). The trackis enclosed in a building 9.8m(30ft) wide and 9.8m(30ft) high and, during operation, thebuilding is tightlysealed to ensure still air along theflightpath.The facility has been equipped witha model support system (Fig.63) thatallows small amounts of model translational motion freedom, aswellasangular motion about a ball-bearing gimbal system. Any motion withrespect to thecarriage is measured and used asanerrorsignal in a closed- loop servomechanism which can position thecarriage withrespect to themodel. If everything is working properly, thereis, essentially, a free-flying model about a mean freeflightpath withwhich to reference the natural dynamics of themodel. Note alsothatthefacility can beused for steady-state (Putnam refers to it asstatic)testing in which themodel moves along thetrackat a constant verticalposition.Aversion of the subsonic tilt nacelle (Grumman 698) V/STOL aircraftwas tested asreported in Ref.78, The principal objec- tive of thetestswas to evaluate transient ground effects. Although theresults have notbeen completely analyzed, experience indicates thatwiththis method of testing, thedata should represent those of steady state. For another configuration, a comparison of lift and drag data is reprinted from Ref.77in Fig.64 (coefficients arebased onslipstream dynamic pressure). The data show good agreement withwind-tunnel test resul_ts.

5. MODEL DEVELOPMENT

The V/STOL wind-tunnel model continues to beone of themost expensive parts of any wind-tunnel program; it will beeven more soif we areto takecomplete advantage of recent advancements in model fabrication techniques and instrumentation. The unpowered high-speed models or those adaptable to pressure tunnels are expensive to contour and finish, particularly if equipped withsurface-pressure taps;thehigh costis a result oftheprecision required and large design dynamic pressure. Adding power can double or triple any model costs because of (I) thecostof theengine simulator, (2)thepower source (electric,compressed air, etc.)hookup, and control,(3)theadded instrumentation needed to evaluate power setting, and (4)theadditional effort required for data reduction. Once theabove complications areadded, there seems to belittle directcorrela- tionwiththesizeand costof themodel itself sothat,in general, withintherestrictions of given available wind-tunnel sizes, scale effectscontinue to dictate aslarge amodel aspossible. Except for possible limi- tationsowing to theavailable propulsion simulators, theinitial approach in model planning is to evaluate themagnitude of wind-tunnel wallconstraints. Toputall of this together requires proper use of material and construction techniques in order to minimize flexibility, change time,or shop costs.The above consider- ations will bediscussed in more detail.

5.1 Model Sizing

Figure 10was proposed asa correlation of experience in comparing full-scale wind-tunnel testresults withthose of actual flight testsusing, in some cases, thesame airframe. Many investigations have, subse- quently, evaluated model sizelimitsand related these limitsto themagnitude of wall-constraint corrections.

Approaches towall-constraint correction might beorganized intoapplication to type of lifting systems as follows: 1. Power-off or low thrust deflection: use classical corrections from Pope (Ref. 6) 2.

Power-on with distributed blowing and limited thrust deflection: use Pope based on CLA 3.

Power-on concentrated or focused thrust deflection: Heyson (see Ref. 79 for summary) 4.

High angle of attack, Maskell (Ref. 80) with considerations by Peitzman (Ref. 81) and Stoll (Ref. 60) 5.

Blockage for all cases (Ref. 82) 6.

Corrections using wall pressure signatures (Ref. 83) The Iast two items have shown promise for reliable evaluation of wall-constraint effects combining simple modeling of the lifting system with wall surface-pressure measurements in the wind-tunnel test section. These corrections increase rapidly with relative model size to wind-tunnel cross-sectional area and will be enhanced by the rapidly developing computational methods, such as was done recently by Snyder and Ericksen at Ames (Ref. 84). In the latter case, PAN-AIR (a high-level paneling code), was used to evaluate wall corrections with emphasis on a bump under the model. The basic problem and flow modeling are shown in Fig. 65. Lower- level paneling codes and vortex-lattice simulations of walls and lifting systems are getting more attention for treatment of the wall-correction problem and are being integrated with the data acquisition software of some wind tunnel facilities.

The question now is at what point do these corrections become meaningless or questionable. The physical constraint for highly deflected jets islclassically illustrated in the sketches of Fig. 66, which are reprinted from Ref. 5, and are taken from the study done by Tyler and Williamson (Ref. 85). Also included is an example of one jet height and several velocity ratios which were studied for several relative test section sizes. The general flow pattern at the limit of testing consists of a vortex surrounding the point of jet impingement on the wind-tunnel floor. In this case (Fig. 66), for two laterally spread jets, the jet exhaust impinged on the floor at V_ = 1.31De/h For the lift-fan data of Hoad and Gentry (Ref. 86), Margason (Ref. 5) runs through an example to keep As at the tail to 5 ° , and using Heyson's corrections, Vemin = 0.125. Testing below this velocity would probably cause a vortex formation. There is a suggestion in Ref. 87 that testing should be limited when the wake impinges 2.5 wing spans downstream from the model. This is consistent with Turner's criteria of Fig. 54 for the moving-belt ground plane.

Carbonaro (Ref. 88) considers limits using Heysons criteria of testing for acceptable amounts of correc- tion. A set of limits of wall corrections was assumed, such as Ai t = ±5 ° , ±2 ° , and ±1/2 ° for maximum acceptable, moderate, and no corrections, respectively, as listed in Table 3. Values of maximum test lift coefficient are related to ratio of wing span to tunnel width for several semispan and complete lifting sys- tems. In this and previous reviews, Carbonaro concluded that the flow breakdown limit applies mainly to models that are small with respect to the test section; that is, for span-to-width ratios below about 1/4.

Typical lift versus test-section width plots are reprinted in Fig. 67. For larger model dimensions relative to test-section size, the wall corrections become too large before flow breakdown occurs.

In the author's opinion, many of these limits can be "pushed" and some valuable information still obtained on power-induced effects for such items as aircraft stability and control, as well as flap and control loadings. Of equal significance is the fact that Heysons early work using a linear wake trajectory has been replaced by a free or "relaxed wake" which tends to move the predicted point of impingement on the wind-tunnel floor downstream. As previously mentioned, low-level CFD methods simulating the complete lifting system as well as the floor are being considered for evaluating both sizing criteria and wind-tunnel wall corrections, including support influence.

The general case of blockage and wall-constraint effects has been studied for some time by Hackett. An updating of this work is published in Ref. 82. The monitoring of wind-tunnel wall pressures and relating the results to an equivalent distribution of sources and sinks continues to be a promising method of evaluating the effective blockage of most conceivable V/STOL aircraft configurations. If properly instrumented, an angle-of-attack correction can also be derived, as illustrated in the block diagrams of Fig. 68(a). The key is thecomparison of themeasured pressure imprint onwalls(Fig.68(b)) withresults of theoretical modeling of thelifting system.

Investigations intomethods of extending these limitshave included changes in wind-tunnel design and methods of mounting themodels. Off-center mounting, such asmoving themodel higher in thetestsection to effectively increase De/h, may help in some cases, butasshown bytheUniversity of Washington studies (Refs. 89and 90)it tended to increase themagnitude of thewall-constraint corrections. Adding slotsto both thefloor and ceilingcould beeffective in delaying theformation of thefloorvortex.Several moder- atelysized wind tunnels areequipped withhardware needed for this. The systems and correction factors may besimilar to those being posed for transonic testsections (Ref. 91). Unfortunately, theamount of suction required through theslotsand thesystem for distributing this suction is still being debated, particularly for thecase of high-velocity jet impingement. Along withthis debate goproposals of so-called adaptable walls. Sears' concept (Ref. 92),would require linkingthewall-constraint porosity and shape to a mathemati- calmodel of thewake in order to eliminate any possibility of thewake being ingested bythelifting system (Fig.69). Small-scale experiments onthis concept arenow under way, butit seems unlikely thatit could be adapted to any of theexisting large wind-tunnel facilities.

Asthesizeof thelifting system increases withrespect to thetunnel test section, theincrease in blockage and,hence, free-stream velocity correction, resultsfrom thecombined effects of model bulk, deflected jets, or separation wakes. Forhigh-angle-of-attack tests,amodel sized for conventional angle-of- attack ranges can have sufficient blockage at angles of attack above 45 ° tomake test results meaningless.

Fortheinvestigation of Refs. 60and 61,a correction was used which was based onflat-platedrag datasuch asthose shown in theplotof Fig.70(a).The gross drag onthemodel and supports was measured bythe tunnel scale system. From this, model-induced drag, lessgross thrustand inletmomentum drag, was subtracted and related to a Aq corection in Fig.70(b).The blockage correction was assumed to actuniformly across thetest section.Asshown in Fig.70(b), theresultingAq correction for themodel with S/C =3.5% was 11% at 90 °. Formodels having much larger values of S/C, theflowuniformity assumption would bein question.

5.2 Power Simulation

In evaluating V/STOL concepts in thewind tunnel using scaled models, it is becoming evident thatthe full-scalejet properties must beevaluated in greater detail. Early experimenalists studied engine-airframe aerodynamic interactions using jets of air thequality of which was seldom documented. Asa result,there were differences between experimental results which could notalways beexplained bysuch things asvariances in velocity ratioand placement Qfthejets.

Ames Research Center has undertaken a major experimental program to evaluate theprincipal scale effects involved in jet simulation. The initial phase, reported onin Ref.93,included thefull-scale testportion of a full-scale to small-scale comparison program. Measurements of theeffects of using actual turbojet/ turbofan engines will beevaluated at small scale byattempting to simulate thegeometry and jet characteris- tics of theengine exhaust and evaluating anyfull-scale differences in suck down and ground effect. The results, using a General Electric YJ-97-GE-100 turbojet engine, aresummarized in Fig.71. Although small- scale experimental results arenotyetavailable for comparison withexisting large-scale data, it is evident thatthemeasured suck down is considerably more than thepredictions using empirical methods (Ref.94)which were based onsmall-scale generic investigations. Thisprogram will continue withthesmall-scale phase and thenbeextended to co-flowing jets, butuntilwehave answers onthesignificance of modeling all jet charac- teristics,including thepossible need to simulate jet turbulence, thecurrent objectives should beto model therealjet properties of jet cross-section and velocity profilefor a given pressure ratio.

The need to dothelatteris demonstrated in theexperiments of Kuhlman and Ousterhout (Ref. 95)and Kuhlman and Warcup (Ref.96). Aplug was inserted in a nozzle to change thejet qualities, particularly the total pressure variation across thejet. Atypicalresultis shown in Fig.72. Inserting theplug near the jet exit added a lossin momentum to thecenter of thejet and contributed to a more rapid velocity decay and change in jet trajectory in thecross flow.

The foregoing results indicate thatsimulating jet velocity profileacross thejet is mandatory and, possibly, thesimulating of jet turbulence could besignificant for evaluating jet-airframe interference effects. Astherequirement for more quantitative testresults increases there is a need for a more careful choice of thepropulsion simulator to beincorporated in themodel design. In planning thetestprogram, this choice is integral to model sizeconsideration because of theavailability of off-the-shelf simulators and the large costof simulator unitsalready developed.

5.2.1 Simulator Classification

Classification of aircraftengines, and hence engine simulation types, was wellorganized byWulf and Melzer (Ref. 97),asshown in Fig.73.

The choice of simulation is between "self-manufactured motionless systems" and "expensive, purchased rotating systems," withrotary-wing model power and turbofan engines falling in this latter class. Although notlisted,hydraulic-powered motors should notbeoverlooked for driving rotary-wing models or for installa- tionsin which a return lead can beinstalled.Table 4 bridges thetwo branches for air-driven models classi- fied according to nacelle type.Toextend theapplication to full or large scale, I added theline item, gas-turbine engines. The following threesections cover thefixedblown simulators which include thefull, blown jet, and ejector,aswellastheturbine powered simulator and gas turbine engines. Electrical power is considered beyond thescope of this lecture though it is a significant factorin model costs. Its application is typicallya mechanical problem and one of tyingintoeitheranon-board model motor or to a wind-tunnel auxiliary power facility, such astheone available at theVertol tunnel (Fig.51)or attheAmes 40by80 (Fig.9).

5.2.2 Fixed Blown Simulators

Fixed blown simulators probably have been and will continue to beused because of theireconomy, particu- larlyfor powering partiallycomplete models. Most initial studies in jet/airframe interference effects have incorporated this type of power fer jets, such as the work by Vogler (Ref. 66) and, more recently, in the jet-in-cross-flow studies of Refs. 44 to 48. The simulation of a uniform jet and, at the same time, getting it within the contours of a small model of a complete aircraft configuration, is a difficult task, particularly if the entire assembly must be metric (total forces measured by the balance). Effective designs of the fixed blown simulator were used in the early studies of the Hawker P-1127 using a I/lO-scale model; an illustration of the model is reprinted in Fig. 74 (Ref. 63). In the most recent development of the AV-8B, a O.15-scale model was powered by the airfeed plenum shown in Fig. 75.

In any of these, and also in similar designs, the quality of the jet is a function of the screen or colander design combined with the size of the settling chamber. A very uniform jet velocity was obtained in the AV-8B model just mentioned. A very compact installation, it still had a settling chamber up stream of the conical screen (Fig. 75). In these installations the desired contraction from screen into the nozzle entrance was difficult to package. For the basic jet-in-cross-flow studies of Aoyagi and Snyder (Ref. 46) and Fearn and Westen (Ref. 47), a very uniform jet velocity profile was obtained by the relatively bulky design shown in Fig. 76. The need for some contraction ratio is always a major problem in designing fixed blown propulsion simulators for complete aircraft models because of limited available space inside the model.

For simulating (approximately) engine inlet velocity, as well as in providing exhaust flow with a minimum of high pressure flow, the ejector should be considered. The application of the ejector was aided by the work published in Ref. 98. The ejector used was fed by air with pressures up to 2.41MN/m 2 (350 psig) through a large number of primary nozzles in order to maintain uniform flow across the ejector. Ejector performance for one design is presented in Fig. 77. Even though an overall pressure ratio of 1.5 could be obtained, the ratio of ejector inlet flow (secondary flow) to total exit flow (primary plus secondary) was less than 0.5, meaning a subscale inlet area. Jet-pump theory indicates that this ratio would increase rapidly as less exit pressure is required. The effort did show that effective gross thrust for any given ejector configuration could be set to a known value and maintained, providing a complete calibration was made for each exit nozzle or deflector- inlet combination. The ejector is adaptable to packaging into nacelle configurations, as shown in Figs. 78(a) and 78(b) for a turbojet or turbofan, respectively (Refs. 97 and 99). Subsequent development such as the effort by General Dynamics (Ref. i00) has refined the use of the ejector for simulating high bypass ratio engines. (See also Ref. 99 for application to an investigation of the EBF- externally blown flap.) The simulator used here, and shown in Fig. 78(b), simulated the correct exit thrust split but the incorrect inlet flow simulation, though probably close enough to have had little effect on measurements of EBF performance.

5.2.3 Turbine Powered Simulators: Low Pressure Ratio The tip-drive fans (Fig. 79) are roughly one tenth the cost of the center turbine driven fans but still have a big advantage over the ejectors in being able to move the bypass flow at 5 to 6 times the flow rate of the drive air. As for U.S. designs, a major workhorse has been the 5.5 in. (14 cm) fan shown (Fig. 80) disassembled and installed in an early Rockwell lifting-nacelle configuration (Ref. I01). The fan was also used to simulate the fan flow in the built up nacelle (Fig. 81 from Ref. 102). The 5.5-in. fan was combined with a direct blown simulation of an engine core in a O.094-scale model of a V/STOL research airplane to be powered by Allison PD 370-16 engines. For this test, the fan turned out to be a reliable performer. One of the largest of the tip-driven type was Tech Developments 12-in. fan (Ref. 103), which was incorporated in the high-angle-of-attack model tested in the Ames 40 by 80 tunnel (Ref. 60) and in ongoing inlet work at NASA Lewis. In both of these cases, it has been used only to suck the inlets rather than provide a propulsive jet since, at this stage of its development, it has a large radial variation in total pressure. Initial calibra- tions established, however, that it met the design objectives of an effective pressure ratio of 1.4. Air supply required for a full total thrust of 800 Ib is 6.5 Ib/sec heated to 250°F. Again, testing experience has shown this fan to be reliable.

In the final stages of aircraft aerodynamic development, the turbofan or central powered simulator has had a strong role. Here, the engine bypass and pressure ratios are both closely simulated for given thrust loadings, T/A L. As a result, for a given nacelle configuration, scaled to the same AL/S as the aircraft, testing can be accomplished at high subsonic Mach numbers for effective evaluation of cruise drag. The TF 34 simulators are also being used in the small-scale tests of the Grumman 698 (subsonic tactical aircraft) model shown in Fig. 82 mounted for hover testing in the Ames 40 by 80. Results have not been published for these tests, but long-range plans involve comparison with full-scale static and wind-tunnel tests, with documenta- tion of the jet efflux coming from both large- and small-scale models.

5.2.4 Compact Multimission Aircraft Propulsion Simulator The ejectors and rotating simulators just discussed can be applied only to simulated turbofan engines having subcritical pressure ratios. For power-on testing of high-performance aircraft, techniques used in the past have required direct feed of the air and, if the inlet were to be simulated at the same time, a bulky suction lead would have to be routed outside the wind tunnel. The continuing development and application of the compact multimission aircraft propulsion simulator, CMAPS (Fig. 83), has had the objective of simulating both propulsive jet and inlet flow at total pressure ratios corresponding to those of current high-performance engines. The simulator (Fig. 84) is a miniature four-stage turbocompressor driven by high-pressure air expanded through a single-stage turbine. It has the capability of changing the engine pressure ratio at a constant compressor airflow by changing the fraction of turbine discharge air of the exhaust nozzle. The hardware development started by McDonnell Aircraft Company under the auspices of the Air Force Aero Propulsion Laboratory; a complete history of its development is provided in Ref. 104.

Ames Research Center is funding a major program to develop the technology for the application of the CMAPS to small-scale wind tunnel models. The primary objective in this program is to measure the aerodynamic interaction effects that may result from geometrically close-coupled propulsion/airframe components. A second objective is the development of installation and test techniques for propulsion-equipped wind-tunnel models.

A third objective is the expansion of the high-speed V/STOL aircraft aerodynamic data base. The initial aircraft configuration chosen was that of a two-engine, close-coupled, canard-controlled aircraft similar to the General Dynamics design (GD 205) resulting from the study of Ref. 2. The model is shown installed in the Ames 11-ft transonic tunnel in Fig. 45, and the basic model design features are described by Bailey et al.

in Ref. 105.

As the model was developed, the problem of isolating the aerodynamic forces acting on the airframe from the propulsion forces was addressed, as shown in Fig. 85. As it turned out, a single internal balance was used which supported all external surfaces except the boattail. The boattail forces were measured using surface- pressure instrumentation. In this manner, the use of seals was kept to a minimum, with one at each inlet and exit of the simulator and at the intersection of the support and lower fuselage.

A major part of the program has been the development of the calibration tank (Ref. 106) along with the control system itself, and the instrumentation needed to monitor inlet flow (Ref. 107). The general purpose of this facility (Fig. 86) is to obtain pretest relationships between the inlet and exit flow rates and pres- sure ratios as functions of the CMAPS control air parameters. As may be seen in Fig. 84, it will meter inlet air supply and exhaust extractors. All the required pressure valve controls and pressure instrumentation will be linked to a digital control console adjacent to the calibration tank.

5.2.5 Aircraft Engines For large subscale or full-scale V/STOL testing, gas turbine engines are needed. It is at this scale that the larger size simulators, such as the 30-cm (12 in.) tip-driven fan, either have limitations in fan pressure ratio or require an excessive amount of compressed air for continuous operation. For direct blowing, most wind-tunnel facilities do not have sufficient compressed air capacity in either amount, pressure ratio, or heat. For large subscale models, the gas turbine engines developed for business jet aircraft adapt well to scales from 0.3 (for light transport aircraft) to 0.7 for tactical fighter designs. For larger models, there is a jump in thrust to the 10,O00-1b class engines. Relative geometric profiles for some of the smaller engines are shown in Fig. 87 (Ref. 4). Ames Research Center has used these engines in several large-scale research investigations in the 40 by 80 tunnel. The engines have seldom been flightworthy, but have been maintained sufficiently to run at or near maximum thrust in most cases. An installation of the JT15D engine (as bypass ratio 3) used in a large-scale upper-surface-blowing investigation is shown in Fig. 88 installed in a boiler-plate nacelle without some of the nozzle fairings installed. One disadvantage of using this size of engine is that the fuel control and starter are usually located at the bottom of the engine, making it some- times difficult to exactly scale the external nacelle contours of an aircraft configuration. However, this gave few problems in simulating the significant USB geometrical parameters needed for the model of Fig. 89 (Ref. 15).

Maximum rated performance is shown in Fig. 90 (Ref. 4) for an assortment of engines and tip-driven lift fans (LF 336 and LF 376). The augmented wing compressor and the Viper compressor are gas-turbine driven pumps which might be too large to be housed in a fuselage of a high-performance fighter model, but they provide cool air at 30 to 50 Ibm/sec at pressure ratios up to 3.5. The J97 General Electric engine has been used in the basic research models with one shown in Fig. 8 (Ref. 108) and has a moderate combination of thrust-to-pressure ratio. NASA "inherited" all of these engines from a military program; they were in good enough condition to serve as reliable power sources. As can be seen from Fig. 90, there are no engines immediately available in the lO,O00-1b class with sufficient pressure ratio to simulate some of the larger power plants now being planned with pressure ratios of 3 to 3.5. This is forcing the model size to nearly full scale for testing V/STOL supersonic tactical aircraft configurations in order to duplicate pressure ratio.

5.3 Model Planning The foregoing discussions have attempted to treat model sizing and propulsion simulators separately.

This was done primarily to allow emphasis on significant factors for each topic such as scale effects or details of the CMAPS. As was the case for using available engines (previous section), in actual model devel- opment, an experimentalist may have to take whatever power simulation is available. Providing it gives him the necessary parameters such as NPR or thrust, with the support equipment in the wind tunnel that is avail- able to him, he will design his model around that particular simulator. The following are a few examples of V/STOL wind-tunnel model design and construction.

An 11% scale model of the Grunlnan 698 tilt nacelle aircraft is currently undergoing wind-tunnel tests in several U.S. facilities. This high-speed steel and aluminum model was designed and built simultaneously with the hover and wind-tunnel testing of the full-scale boiler-plated model. From previous development of conven- tional military aircraft using the TF34 simulator, sufficient experience with the unit indicated that it could be a reliable power plant for a V/STOL model. The model was constructed with provisions for both three inter- nal balances and 200 surface-pressure taps, which resulted in the tubing network (Fig. 91) leading to the on-board Scanivalves. The nacelles, shown partly disassembled in Fig. 92, were designed to house the TF34 simulator, supply the required amount of drive air, and attach to the model through a floating or metric frame so that the propulsive forces and moments could be measured independently of those acting on the wing fuselage.

In the process, a coil (Fig. 93) for the fan drive air was developed to minimize force and moment tares from the air. This was done before the air entered the model since there was not enough room inside the model to house isolation systems such as flow-through balances (see Sec. 6.1). As might be expected from the complexity of the model, there were several problems that had to be addressed, not the least of which was the excessive time needed to assemble the model on the sting. Tests using the model have been considered successes since reliable data were finally obtained and are now being analyzed. The final cost of this model corrected for 1984 dollars was probably over $800,000, and this did not include the cost of the TF34 simulators (approxi- mately $100,000 per unit).

The prime question is could the above model have been designed in such a manner that the cost would have been lower. The answer is that it probably could not have been, although there were alternative designs which might have included two complete models, one being a force model with no surface-pressure instrumentation and the other strictly a pressure model. This is commonly done for high-speed conventional models, such as that shown in Fig. 94, which shows the model plus some of its components needed for alternative configurations and model changes. However, the principal cost of such models comes from the close tolerances required to maintain scaled contours and model components that are interchangeable. Rapid advances in computer aided design and manufacture (CAD-CAM) techniques where applied to these complicated wind-tunnel models is reducing some of the time for construction, but there will always be a need for a certain amount of hand fitting, which runs up costs. Another option for the tilt-nacelle model, could have been the use of an unpowered or transi- tion test that if the proper propulsion simulator had been available, could have been of larger scale but sized (about 0.2 scale) to the limits of testing in a moderately sized subsonic tunnel, such as the Langley V/STOL tunnel. The actual contouring tolerances could have been relaxed and wood-plus-foam-fiberglass mate- rials considered. Even so, if the on-board pressure instrumentation were needed in both high- and low-speed models, the total costs of both models would be the same or more than the one that was built.

For large-scaled V/STOL models designed for testing in the NASA large-scale wind tunnels, the options in model design and construction are more numerous. Levels of sophistication and particular examples are shown in Table 6 defining several levels of costs. Level I would include the component tests such as the Q-fan (Fig. 32) or the high-angle-of-attack studies using the O.4-scale model of Fig. 48. Level 2 would include large but probably subscale models of mostly boiler-plate construction and probably powered by the smaller gas turbine or turbofan engines mentioned in the previous section. Level 3 would include full-scale, mostly boiler-plate models using the same or similar power plants as planned for the aircraft, such as the AV-8B flight-like model of Fig. 5. Since this type of model is full scale, it could incorporate flight-weight com- ponents where appropriate. Level 4 would be wind-tunnel tests of the actual aircraft remotely controlled as required and adapted to the wind-tunnel support struts, using either the landing gear attachment points or reinforced wing-mounting arrangements. The increasing cost with level applies principally to tactical V/STOL aircraft for level 3 and below since the actual aircraft might be wind-tunnel tested cheaper than the total costs of building and testing a large but subscale fully powered model.

The term boiler plating might wrongfully have the connotation of making something overly heavy and strong with little regard for model weight and some of the aircraft details to be modeled. This is generally not the case at all though some models turn out heavy. What does reduce construction costs is to stay away from com- plex surface curvatures by using straight-line elements and rolling or multiple bending of the skin. Care must be taken in the planning of the model so that short straight-line elements simulate the scaled, gradually changing, surface contours. Stamped or machined elements are used for sharp curves such as leading edges or inlets. An example was the USB STOL model (Fig. 89) which incorporated the Pratt and Whitney JTI5D powered nacelle of Fig. 88. For this model, the wing leading edges and flap skin were formed using multiple bends along straight-line wing elements and the inlets were spun-formed for blending with the skin of the nacelle.

Most of the rest of the model was boiler plated, using welded steel construction, as shown in Fig. 95(a) for the two-engine fighter model (Fig. 8, Ref. 108). Where the surfaces are not exposed to hot flow, a rigid and complex surface is obtained by polyurethane foam, which is applied in liquid form between plywood ribs, allowed to expand, and then cured. This is then shaped and covered with several layers of fiberglass and filled.

Going one step further, this technique can be incorporated with aircraft parts such as was done for the full- scale AV-8B model shown in Fig. 96 (Ref. 19).

6. TEST OPERATIONS Planning and carrying out the actual wind-tunnel testing operation must depend not only on the capability and test limitations of the model but must also rely on full use of support and equipment available at the wind-tunnel facility. This is particularly true for the powered V/STOL model, testing at all speeds, where such things as time to set power, tunnel venting (where actual engines are used), change time, or time to take a data point can determine the success or failure of the test operation. Examples of the latter are factors considered in Ref. 4 and reprinted in Fig. 97. The actual data were obtained from experiences with gas-turbine powered models in the Ames 40 by 80, but they are significant factors to be dealt with for any powered-lift testing, particularly, downtime for configurational changes and time per data point. Many other factors influence facility use time. The ones chosen for the following brief discussion are instrumentation and data acquisition, acoustics studies, and flow surveying.

6.1 Instrumentation and Data Acquisition It will be assumed that the experimentalist has available to him the most up-to-date instrumentation with potential for on-board use such as Scanivalves (if possible to install on,the model) or a reliable flow-through balance similar to the one shown in Fig. 98, which was used in the investigation of Ref. 109. For power coming from cool compressed air, this design integrated the force isolation scheme for the drive air with the balance itself. In this case, opposing bellows and seals were used with a metric mass-flow distributor routing the air to various blowing systems in the model. When the air must be heated above about 200°C, the flow-through feature might not be practical, a result of thermal sensitivity of the force-measuring part of the balance. In such a case, drive air might have to be "jumped" through an isolation coil similar to that shown in Fig. 93.

As for Scanivalves, at the date of this lecture, many test facilities are converting to the electronic type which takes advantage of low-cost silicon pressure transducers whose inherent errors (such as thermal zero shift and output drift with time) are corrected by the periodic on-line calibrations. The advantage of this system, particularly for a powered model testing under fluctuating load, is that all pressures can be sampled almost simultaneously within the channel capacity of the data-acquisition system. At first glance, some of the possible installations seem geometrically smaller than the smallest of the mechanically driven ones (although this may not generally be the case). A possible disadvantage for low-speed V/STOL use is a measure- ment error resulting from the requirement to use a ±5-1b/in. 2 transducer for a ±l-lb/in. 2 measurement range.

A unit of this range, however, will be available shortly. Other uncertainties are possible adverse effects caused by temperature and by high-frequency vibration of the mount.

For large-scale testing, an on-board data-acquisition system using pulse-code and modulation (PCM) similar to that used in flight testing, has been used for several years with the remote digitizer multiplexer unit (RMDU) which is evolving into the remote millivolt multiplexer and amplifier module (RAMM) (Ref. 110). This allows all conditioned analog signals except those from thermocouples to be digitized and multiplexed so that they can be transmitted to the data acquisition through a minimum of leads. The compatibility of electronic scanners with this system is under evaluation.

The wind-tunnel data-acquisition system thatis finding acceptance at Ames Research Center withsimilar systems in industry was detailed byCambra and Tolariin themid-1970s (Refs. 111 and 112).The version finally adapted for theAmes Large Scale TestFacilities (40by80)is shown in Fig.99. It is sufficiently flexibleto beapplied to both dynamic and steady-state testing. Alsoshown aretheother operating systems, including theinputs from themechanical (below thefloor)wind-tunnel balance (Toledo system) which aretrans- mitted to thedata gathering processor (DGP) through thecontrol processor. The real-time executive processor (REP) processes information from theDGP and rapidlyreturns theresultto thedisplays in thecontrol room.

Fornormal V/STOL testing, real-time updating of thedisplays, which might include aerodynamic and propulsion parameters, every I sec has been found satisfactory. The existing choice in software provides a large range in timefor data sampling during therecording process. Fordynamic testssuch asrotary-wing investigations, thedynamic analysis system (DAS) is available to beused eitherseparately or combined withtheDGP.

6.2 Special Requirements for Acoustic Studies

The strength of all possible noise sources onanaircraftshould bedetermined during thefinal develop- ment of theaircraftin order to obtain theacoustic loading and to ensure a safe working environment for both thecrew of theaircraftand ground-support personnel. FortheUSB testsof Schoenster etal. (Ref. 113) there were small butnoticeable differences in surface acoustic loads owing to airspeed even though theairspeed tested (in theLangley 30by60)was only16m/sec. The investigation of the727/JT8D flight noise (Ref. 114) compared model, full-scale wind-tunnel data, static-test data, and levelflyoverdata for the727 airplane.

The JT8D installation shown in Fig.100 was used withmicrophones ona traversing beam 3 mto therightof the engine centerline (looking aft) butstill in thenear-field.In thisway, theangle relative to thenoise source could bechanged. When themeasurements were extrapolated to thefar-field,they compared wellwiththe aircraftforward-speed effects, asshown in Fig.101.The investigation has been one of several which pointed outtheneed for and value of continued acoustic study in thewind tunnel.

6.2.1 Method

Problems withand techniques for measuring strengths of noise caused byjets and inletsin thewind tunnel were evaluated bySoderman and others in the1970s (Refs. 115-117). Measurements of noise radiation from powered V/STOL models can, generally, beconfined to using incomplete or component models of theprimary noise systems aswas done for theBoeing 727 study just mentioned. Inthisway, thestudy of forward-speed effects ontheacoustics of anaircraftcan beinvestigated at full-scale withaccurate model simulation of theantici- pated noise sources. Anexception to thiswas theevaluation of airframe noise (Ref.118) where thenoise radiating from a large-scale semispan model of thecomplete lifting system was measured in the40by80.

The principal problem areas in measuring noise in awind tunnel aresource-noise reverberation, background noise owing to thewind-tunnel drivesystem, and microphone wind noise.Toalleviate these, some tunnel facil- ities areupdating theequipment and, in effect,doing thefollowing:(1) acoustically treating thewalls, (2)reducing fan-drive noise, and (3)reducing wind noise byrefiningsupport strutsor eliminating protuber- ances from thewalls. The challenge is one of selecting measuring devices thataredirectional or discriminate against unwanted sound, and byusing special techniques and free-fieldcomparisons to evaluate thereverberant field. An example of thelatter is reverberant field simulation in which thereverberant sound field of com- pact wind-tunnel models is simulated byoperating acalibrated loud speaker at thenoise source location (Ref.119).When thereverberant field levelis known, themodel noise data can becorrected to giveapproxi- mate free-field noise levels. The technique was verified bywind-tunnel and flight correlations using the actual flyover noise, such astheXV-5 comparison of Fig.102 (Ref. 120).Unfortunately, where thenoise source is large and distributed relative to thesizeof thetunnel testsection, such asfor spanwise distrib- uted blowing models, thetechnique cannot beapplied, and reverberation corrections must come from operating both in thewind tunnel (at V_=O)and in thefreefield.

6.2.2 Forward Velocity Effects onAdvanced Inlet Suppression of Fan Noise

The inlet-fan acoustic studies of Falarski and Moore (Ref.121) and of Moore (Ref.122) areother examples in which thewind tunnel was effective in evaluating forward-speed effects onacoustic sources. Flighteffects onfannoise have been observed byinvestigators who have compared turbofan flyover noise withstaticnoise, butthedetailsareobscured bythemixture of aircraft-noise and other engine noise sources. Therefore, in order to understand fannoise, theactual flight conditions must besimulated. Tosuppress compressor noise, advanced inlet designs were devised. One design, thehybrid inlet, had a smaller than normal throat area and acoustic treatment in thediffuser. The other, thedeflector inlet, had anextended lowinlet lip withacous- tic treatment to deflect fannoise upaswellasforward. The performance and acoustics of both of these inletswere to becompared withthose of a standard inletwithnodiffusion.The objectives of theprogram were to determine thelow-speed flight effects assimulated bythe40by80ontheforward radiated fannoise and ontheacoustic suppression characteristics of theinlets.

The series of testinletswas designed around therequirements of theJTI5D, withtheresulting basic nacelle shown in Fig.103.The turbine and thefanexhaust ducts, aswellastheskinof thenacelle, were acoustically treated.The simulated flight testsin theAmes 40by80used thetestinstallation shown in Fig.24(a), which is shown withmicrophone stands installed and witha circular traversing stand mounted in frontof themodel. The entirefan-strut assembly was mounted ontheturntable of thewind tunnel for angle- of-attack adjustment. In both installations, a foam mat was installed which removed reverberant reflections from thenoise data at all frequencies above 500 Hz. Asa result,thefan-nacelle installation, using the JT15D asmounted was about I/4-scale of moderately sized current, commercial turbofan aircraft'sengines. To enhance simulation of thelarger engine, thecore IGV (inletguide vanes) had been modified. The traversing microphone was kept aligned withthefreestream byusing a vane-type pivoted mount. The standard 0.25-in.

microphone was used witha fairednose cone attached to reduce wind noise.Noise inside theinletwas mea- sured using Kulitepressure transducers (outside testsonly)mounted flushto thelocalsurface.

Care withthedetails justmentioned, aswellaswithmany otherfactors thatinfluenced t_edata, even- tuallyproduced some interesting results, the foremost being that the hybrid inlet suppressed the high-tip- speed fan noise as much as 18 PNdB on a 61-m (200-ft) side line which was scaled to CF6 size engines. In addition, it was found thatnosignificant changes in fan-noise suppression for eitherinletoccurred for forward-velocity changes above 21m/sec (68ft/sec)or for angle-of-attack changes upto 15 °.

Thisis another clearinstance of wind-tunnel acoustic studies helping in theevaluation of flyover noise.In another sense, it verifiedsome of thenoise evaluation techniques prevailing today and which are applicable to acoustic studies of V/STOL aircraft.

6.3 Flow Surveying

ForV/STOL wind-tunnel testing operations, flowvisualization, if notflow-field surveys, can bea major partof thetestprogram. The traditional method of using surface tufts, tuft or smoke probes, and directional flowsurveying arenecessary toolsin understanding thecomplicated flowpatterns which surround a powered- lifting system, and how they influence theaerodynamic parameters of theaircraft. Forexample, in thestudy and modeling of lifting jets, theflowvisualization testsof Margason (Ref. 123) have been considered bysome to besufficiently quantitative to define thepath of thejet-in-cross flow.

6.3.1 Surface Effects Study

No subsonic test of a lifting system is complete without themodel being covered withtufts, oil, or kerosene. Studies such asthatshown in Fig.104 areinstrumental in evaluating flap performance (and, in this case, model installation effects). For a review of some of these techniques used near themodel surface the reader is referred toWerle (Ref. 124) (ageneral treatment of all flow-visualization technique) and Merzkirch (Crowder) (Ref.125), who has specialized in theuse of fluorescent minitufts for nonintrusive surface-flow visualization.

In addition to qualitative information thatflowvisualization has always provided, more quantitative information is now being obtained, from theimaging and numerical processing of flow-visualization pictures (Ref.126).Fortwo-dimensional flow,flowseparation is becoming wellunderstood, butin three dimensions theproblem still needs considerable research. Forinterpreting flow-visualization datain three dimensions, thecurrent trendis to use topological concepts withskin-friction lines. Although these concepts have been withusfor some time,Hunt et al. (Ref.127) initiatedmore interest in thesubject.Recently, Kao and Burstadt (Ref.128) applied themethod to analysis of visualization data for deflected thrustV/STOL nozzles, and others have applied it to flowseparation at high speed. More application to analysis of flowpatterns on V/STOL aircraftin hover and transition should becontinued in view of theeconomy ofobtaining this type of data.

6.3.2 Flow-Field Measuring

Methods incorporating pressure probes, LDV, IRimaging, Schleiren, shadowgraphing, and hot-wire surveying arebecoming wellestablished. The use of smoke is becoming more common for evaluating flowfields,and, its use and limitations arethoroughly discussed byWerle (Ref.124) and Mueller (Ref.129).Where a quick approx- imate answer is required, theclassical tuft gridshould notbeforgotten. The results of Naeseth (Ref.130) were certainly descriptive of theeffectof high power induced circulatory lift ontheupwash near theengine inlet (Fig.105).In view of theneed of survey equipment and methods thatcan beused for three-dimensional flowin a hostile environment, such asis thecase for V/STOL testing,thefirst two,pressure probes and LDV, will bediscussed. Where thereis a need for close study of theflowfield involving high-energy jets, partic- ularlyat small scale and high speed, a special attempt should bemade to utilize thewind-tunnel Schleiren and shadowgraphing equipment.

The use of pressure probes and rakes has been essential throughout thehistory of experimental aerodynam- ics. Rake survey equipment, boundary-layer rake probes, and directional pitot tubes have been standard equip- ment in most wind-tunnel facilities. Forevaluating highly deflected flows, thedirectional pitot tubehas continued to beused where more sophisticated methods, such asLDV, arenotavailable. The probes have gener- ally been lacking in three main areas:(1)theirpossible influence ontheflowthatis being measured, (2) lackof sensitivity for small-flow velocities, and (3)errorsin measurement at high angles withrespect to theaxisof theprobe.The first problem, has been reduced- butwill never beeliminated- bytheuse of very small tubing withoff-the-shelf probes of 1/8Dor lessnow being available. The second difficultycan onlybe helped withtheuse of careful calibration and sensitive pressure transducers thataregeometrically close to thehead of theprobe.The third problem has been alleviated recently in two ways.First, theuse of the five-hole probe (Fig.106) (Ref. 47)was enhanced at thehigh-incidence range byusing a potential-flow model to define theform of a calibrating equation. Constants for theequation were then obtained during calibra- tion. Second, theother area has been thedevelopment of theseven-hole cone shown in Fig.106 (Ref.131), which has been calibrated over a large speed range withgood accuracy upto 80 ° incidence.

Laser Doppler velocimetry (LDV) has been advancing steadily since theearly1970s. AtAmes Research Center, Orloffet al. have been extending themeasurement capability from theoriginal two-color-two- dimensional backscatter laser velocimeter (2D-LV, 7 by10installation) to a 7 byI0 sizethree-dimensional laserfacility and a large scale two-component unit. Because of thecapability of obtaining flowvelocity and direction measurements withnophysical interference withtheflowitself, thedevelopment of theequipment will continue. In a recent paper, Orloffet al. (Ref.132) reviewed Ames LVD experience, and plans for and the accuracy of LDV systems arediscussed in Ref.133.

Diagrams of the7 by10three-dimensional laserequipment arepresented in Fig.107.It measures three velocity components bymeans of three independent dual-scatter channels thatoperate in thebackscatter direc- tion. Acombination of mechanical tilt and variable focusing areused to ensure thatthefocalpointof the upper channel remains colocated withthefocalpoint of thelower two-dimensional portion asthelower focal point moves in a cross-flow direction across thetest section during a survey. Streamwise and verticalposi- tioning is accomplished bymoving theentirepackage (Fig.107(a)) onthedigitallycontrolled translation platform of Fig.107(b).The status of this equipment gives repeatable positioning accuracy to lessthan 0.5mm and theprobe has a focalvolume of 1.5mm.It is generally operated in a closed-loop mode withthe computer performing test-point positioning and path control through communication withfive stepping-motor

controllers.Rather thanuse forward scatter,such astheJPL equipment (Ref.134), to increase

signal-to-noise levelfor some testing, thesignal is enhanced byminimizing thebandwidth, theminimum being controlled byprogrammable frequency synthesizers. The equipment has been used extensively in theAmes jet- in-cross-flow program, Fig.108 (Ref.133), where surveys veryclose to theflat plateand nozzle have been completed.

The large-scale laserwas developed to mount in thelarge wind tunnels, asshown in Fig.109 (Ref. 135).

It is a single color,dual-beam, scanning, confocal back scatter LDV thatdirectlysenses two orthogonal com- ponents of velocity; it is shown in operation in Fig.II0. Athirdcomponent is obtained byfurther transfor- mation of coupled velocity components. Asshown in Figs.110 and 111, thelaserand system optics aremounted and enclosed in a streamline cylindrical shellthatis supported along thelongitudinal axis. The laserpower unit and other components aremounted below thecylinder in a fairedstructure which alsoserves asits sup- port. Thisentireassembly is mounted ona carriage for cross-flow translation. The current options for range are10or 20m. This unitwill alsobeused at theAmes Outside Aerodynamic Research Facility(OARF) to study ground-effect flowfieldsand ingestion problems being investigated using large-scale models.

7. CONCLUDING REMARKS

Aspects ofwind-tunnel testing of V/STOL aircraftconfigurations have been reviewed. The choice of topics discussed were, generally, based ontheirsignificance to theplanning stage of awind-tunnel testing program.

Suggestions and comments ontechniques to beused during a testprogram covered a large range in types of test- ingand possible program costs.Their value would begreatly enhanced astheplanning phase of a program matures byobtaining details from thesources listedand, most importantly, thorough consultation withthe staffsof thewind tunnels thatarebeing considered.

The material was organized intothegiven topicsasa resultof technical consideration withlittle, though some, discussion treating theeconomics of testing. Because of ever-increasing costof wind-tunnel time and model construction, wearealways looking for simpler and cheaper ways to construct models and for ways to reduce tunnel use, for example, byfastersampling of thedatafor each testcondition, reduced configurational change time,or reducing thenumber of datapoints byimproved data analysis methods which incorporate use of advanced prediction methods.

The accelerated use of CFD for isolated multienergy fluid dynamic problems which can apply to powered-lift aircraftconcepts combined witha more thorough understanding of scale effects may eventually allow usto feel more confident withhover and transition testing at small scale or models sized for the7 byI0 sizetunnel.

This model would then berestricted in on-board instrumentation. Use of themoderately sized wind tunnels, such astheDNW larger testsection or theNASA Langley 4 by7mrequires a model of a complete aircraftcon- figuration which is currently pushing thestateof theart in miniaturizing on-board instrumentation and power simulators. In addition, even at this scale foratmospheric wind tunnels there arescale effects, and problems in power simulation which might produce uncertainty in theresults. One is therefore ledto consider theuse of large- or full-scale models powered bygas turbine engines and tested in NASA's large-scale testfacilities.

Inthese facilities thereis theopportunity to testwithboth flight-likeand flight-weight hardware, and, for some types of tests,thelevelof total program cost may notbemuch above small-scale tunnel testing.

Although most of my experience has been withlarge-scale methods, I will bethefirst to concede the value of small experiments, notonlyfor exploratory use butsometimes for complete development programs in the design phase. There is ample experience in programs, such astheAV-8B development to show thatit is desir- abletomake good use of small-scale testing in evaluating many of thebasic aerodynamic problems and to verify theresults using large-scale tunnel tests. During anystage of theaerodynamic development, it is recommended thatlarge-scale testing withflight-likemodels bephased intotheV/STOL wind-tunnel testing program to establish hover and transition characteristics. And, finally (formedium sized aircraft),theaircraftitself should betested in one of thelarge-scale wind tunnels, eitherbefore or during flight tests,such aswas done for theXV-15 tilt rotoror theXFV-5 lift fanaircraft. In thisway, aerodynamic predictions, based on small-scale wind-tunnel data enhanced bycomputational techniques can becontinually evaluated.

REFERENCES

1. Hickey, David H.,"VSTOL Aerodynamics: AReview of theTechnology." AGARD CP-143, V/STOL Aerodynamics, 1974, pp.1-1through 1-13.

2. Nelms, W.P., and Durston, D.A., "Preliminary Aerodynamic Characteristics of Several Advanced V/STOL Fighter/Attack AircraftConcepts." NASA TM-81281, 1980.

3. Roberts, L., Deckert, W.,and Hickey, D., "Recent Progress in V/STOL AircraftTechnology. NASA TM-81281, 1981.

4. Koenig, D.G.,Aiken, T. N.,and Aoyagi, K., "Large Scale V/STOL Testing. AIAA Paper 77-586, 1977.

5. Margason, R.J., "JetV/STOL Wind Tunnel Simulation and Ground Plane Effects."Paper No.15,AGARD Conference Proceeding No. 308, Lisbon, Nov. 1981.

6. Pope, A., and Harper, J. J., "Low Speed Wind Tunnel Festing." John Wiley &Sons, Inc., New York, 1966.

7. Pope, A., and Goin, K.L., "High Speed Wind Tunnel Testing."John Wiley &Sons, Inc., New York, 1965.

8. Johnson, D.B., Lacey, T. R.,and Vodo, J. J., "Powered Wind Tunnel Testing of theAV-8B: AStraightfor- ward Approach Pays Off." AIAA Paper 79-0333, New Orleans, La.,Jan. 1979.

9. Lacey, T. R.,and Miller,K., "The AV-8B Wing Aerodynamic Concept and Design." AIAA Paper 77-607, 1977.

2O

10.

Clapper, W.S., Mani, R.J., Stringas, E.J., and Banerian, G.,"Development of a Technique for Inflight JetNoise Simulation. PartI." J. Aircraft,Vol.15,No. 2, Feb. 1978.

11.

Pirrello,C.J., Hardin, R.D.,Heckart, M.V., and Brown, K.R.,"AnInventory of Aeronautical Ground Facilities." Vol.I. Aeronautical Facilities.Vol.II. Airbreathing Engine Test Facilities. NASA CR-1874 and 1875, Nov. 1971.

12.

Aiken, T. N.,Falarski, M.D.,and Koenig, D.G.,"Aerodynamic Characteristics of a Large Scale Semispan Model witha Swept-Wing and anAugmented JetFlap withHypermixing Nozzles." NASA TM-73236, 1979.

13.

Koenig, D.G.,Stoll, F.,and Aoyagi, K., Application of Thrusting Ejectors to Tactical AircraftHaving Vertical Lift and Short Field Capability." AIAA Paper 81-2629, 1981.

14.

Aoyagi, K., Falarski, M.D.,and Koenig, D.G.,"Wind Tunnel Investigation of a Large-Scale Upper Surface Blown-Flap Transport Model Having Two Engines." NASA TM X-62,296, 1973.

15.

Koenig, D.G.,and Aoyagi, K., "Maximum Lift of Upper Surface Blowing STOL AircraftwithSwept Wings."

AIAA Paper 75-868, Hartford, Conn., 1975.

16.

Gambucci, B.J., Aoyagi, K., and Rolls,L. S., "Wind Tunnel Investigation of a Large-Scale Model of a Lift/Cruise Fan V/STOL Aircraft." NASA TM X-73,139, 1976.

17.

Gambucci, B. J., Aoyagi, K., and Rolls,L. S., "Wind Tunnel Investigation of a Large-Scale Model of a Lift/Cruise Fan V/STOL AircraftwithExtended Lift/Cruise Nacelles." NASA TM X-73,164, 1976.

18.

"Wind Tunnel and Ground StaticInvestigation of a Large Scale Model of a Lift/Cruise Fan V/STOL Aircraft."

NASA CR-137,916, 1976.

19.

Hollingsworth, E.G.,Aiken, T. N.,and Ragget, J., "Validation of AV-8B V/STOL Characteristics of Full- Scale Staticand Wind-Tunnel Tests."AIAA Paper 77-597, Moffett Field,Calif., 1977.

20.

Steinle, F., and Stanewsky, E., "Wind Tunnel Flow Quality and Data Accuracy Requirements." AGARD-AR=184, Nov. 1982.

21. Mobey, D.G.,"Flow Unsteadiness and Model Vibration in Wind Tunnels at Subsonic and Transonic Speeds."

C.P. No.1155, BritishA.R.C., 1971.

22.

Paterson, Robert W.,Vogt, Paul G.,and Foley, William M.,"Design and Development of theUnited Aircraft Research Laboratories Acoustic Research Tunnel." J. Aircraft,Vol.10,No. 7, July1973, pp.427-433.

23.

Seidel, M.,and Jaarsma, F., "The German-Dutch Low Speed Wind Tunnel DNW." AeronaUt. J., RAS, Apr.1978.

24.

Duits Nederlands Windtunnel Staff,Noordoostpolder, The Netherlands - Compilation of Calibration Data of the German-Dutch Wind Tunnel. MP-82.01, Mar. 1982.

25.

Herkes, W. H., and Strout, F. G., "Acoustic Evaluation of the DNW Free Jet Shear Layer Correction Using a Model Jet." AIAA Paper 83-757, 1983.

26.

Christophe, J., "The ONERA Wind Tunnels at Mondane Centre and Le Eouga Centre and Their Utilization in Subsonic Range." Office National d'Etudes et de Recherches A_rospatiales, T.P. n 1983-28. Paper was also presented at the 19th Subsonic Aerodynamic Testing Association, Meeting College Station, Kansas, April 18-20, 1983.

27.

Applin, Z. T., "Flow Improvement in the Circuit of the Langley 4- by 7-Meter Tunnel." NASA TM-85662, 1983.

28. "The Boeing V/STOL Wind Tunnel Users Manual." Boeing Vertol Company, Philadelphia, Pa., Oct. 1982.

29. Mort, Kenneth W., Soderman, Paul T., and Eckert, William T., "Improving Large-Scale Testing Capability by Modifying the 40- by 80-Foot Wind Tunnel." AIAA Paper 77-587, 1977.

30. "Guide for Planning Investigations in the Ames 40- by 80-Foot Wind Tunnel Operated by the Large-Scale Aerodynamic Branch." Ames Research Center, Moffett Field, Ca., June 1978.

31. Grafton, S. B., Gilbert, W. P., Croom, M. A., and Murr, D. G., "High-Angle-of-Attack Characteristics of a Forward-Swept Wing Fighter Configuration." AIAA Paper 82-1322, 1982.

32. Phelps, A. E., III, "Summary of Low Speed Aerodynamic Characteristics of Upper-Surface-Blown Jet-Flap Configurations." Paper No. 4, NASA SP-406_ 1976.

33. " The 9M V/STOL Wind Tunnel - A Brief Description and Photo Review of Projects." National Research Council of Canada, Ottawa, Canada, July, 1979.

34. Knott, P. G., "V/STOL Aerodynamic Testing Techniques at British Aircraft Corporation." AIAA Paper 77-584, 1977.

35.

Grief, R. K., Kelly, M. W., and Tolhurst, W. H., "Wind-Tunnel Tests of a Circular Wing with an Annular Nozzle in Proximity to the Ground." NASA TN D-317, 1960.

26.

East, L. F., "The Measurement of Ground Effect Using a Fixed Ground Board in a Wind Tunnel." ARC, R & M 3689, 1972.

37. Betzina, M.D.,and Kita,R.D.,"Aerodynamics Effects of anAttitude Control Vane onaTilt Nacelle V/STOL Propulsion System." AIAA Paper 79-1855, 1979.

38. Demers, W. J., Metzger, F. B., Smith, L. M.,and Wainauski, H.S., "Testing of theHamilton Standard Q-Fan Demonstrator (Lycoming T55-L-IIA Core Engine." NASA CR-121,265, 1973.

39. Betzina, M.D.,and Falarski, M.D.,"Aerodynamics of a Tilt-Nacelle V/STOL Propulsion System." NASA TM-78606, 1979.

40. Weiberg, J. A., and Maisel, M.D.,"Wind Tunnel Tests of theXV-15 Tilt Rotor Aircraft." NASA TM-81177 (also AVRADCOM TR80-A03, April1980).

41. Schoers, L. G.,"Dynamic Structural Aeroelastic StabilityTesting of theXV-15 Tilt Rotor Research Air- craft." NASA TM-84293 (also USAAVRADCOM 82-A-17), 1982.

42. Smeltzer, D.,Nelms, W.,and Williams, T., "Airframe Effects ona Top-Mounted Inlet System for V/STOL Fighter Aircraft." AIAA Paper 81-2631, 1981.

43. Durston, D.A., and Smeltzer, D.B., "Inletand Airframe Compatibility for a V/STOL Fighter/Attack Air- craftwithTop-Mounted Inlets." NASA TM-84252.

44. Schetz, J. A., and Jakubowski, A. K., "Surface Pressured Induced ona FlatPlate withIn-Line and Side-by- Side Dual JetConfigurations." AIAA Paper 83-1849, 1983.

45. Schetz, J. A., Jakubowski, A. K., and Aoyagi, K., "JetTrajectories and Surface Pressures Induced ona Body of Revolution withVarious Dual JetConfigurations." AIAA Paper 83-0080, 1983.

46. Aoyagi, K., and Snyder, P.K., "Experimental Investigation of a JetInclined to Subsonic Crossflow."

AIAA Paper 81-2610, 1981.

47. Fearn, R.L., and Weston, K.P., "Induced Velocity Fieldof a Jetin a Cross Flow."NASA TP-I087, 1978.

48. Fearn, R.L., and Weston, R.P., "Velocity Fieldof a Round Jetin aCross Flow for Various JetInjection Angles and Velocity Ratios."NASA TP-1506, 1979.

49. Mavriplis, F., and Gilmore, D.,"Investigation of Externally Blown Flap AirfoilswithLeading Edge Devices and Slotted Flaps." AGARD V/STOL Aerodynamics, No.143, Paper No. 7, April1971.

50. Aiken, Thomas N.,"Aerodynamic and Noise Measurements ona Quasi-Two-Dimensional Augmentor Wing Model withLobe-Type Nozzles." NASA TM X-62,237, 1973.

51. Riley,D.R.,"Wind Tunnel Investigation and Analysis of theEffects of End Plates ontheAerodynamic Characteristics of anUpswept Wing." NACA TN-2440, 1951.

52. Mahal, A.S., and Gilchrist, I. J., "Design Integration and Noise Studies for JetSTOL Aircraft- Task VIIB Augmentor Wing Cruise Blowing Valveless System." NASA CR-I14,560, Jan. 1973.

53. Anscombe, A., and Williams, J., "Some Comments onHigh-Lift Testing in Wind Tunnels withParticular Refer- ence to Jet-Blowing Models." AGARD Report 63,Aug. 1956.

54. Schoen, A.H.,Kolesar, C.E., and Schaeffer, E.G.,"Static,Noise, and Transition Tests of a Combined- Surface-Blowing V/STOL Lift/Propulsion System." NASA CR-151,954, April1977.

55. "Advancement of Proprotor Technology Task II - Wind Tunnel Test Results." NASA CR-I14,363, 1971.

56. Durston, D.A., and Schreiner, J. A., "High-Angle of Attack Aerodynamics of a Strake-Canard-Wing V/STOL Fighter Configuration." AIAA Paper 83-2510, Oct.1983.

57. Bailey, R.0., Smith, S.C.,and Gustie, J. B., "Propulsion Simulation Test Technique for V/STOL Configu- rations."Society of Automoti:ve Engineers TP-83-1427, Oct.1983.

58. Mraz, M.R.,and Hiley,P. E., "Propulsion Airframe Aerodynamics Interactions of Supersonic V/STOL Con- figurations, Phase I." McDonnell Douglas MDC A7238, St. Louis, Mo., NASA Contract NAS2-10791, July1981.

59. Schnell, W. C.,Ordonez, G.W,and Smeltzer, D.B., "Axisymmetric and Non-Axisymmetric Exhaust-Jet- Induced Effects onaV/STOL Vehicle Desigh. PartIII. Experimental Technique." NASA CR-166,147, 1981.

60. Stoll, F., "Large-Scale Wind Tunnel Tests of a Sting Supported V/STOL Fighter Model at High Angles-of- Attack."AIAA Paper 81-2621, 1981.

61. Stoll, F., Koenig, D.G.,"Large-Scale Wind Tunnel Investigation of a Close-Coupled Canard-Delta Wing Fighter Model through High Angles-of-Attack." AIAA Paper 83-2554, 1983.

62. Laub, Georgene H.,"Low Speed Wind Tunnel Tests ona One-Seventh Scale Model of theH.126 JetFlap Air- craft." NASA TM X-62,433, 1975.

63. Trebble, W. J. G.,"Technique for theAerodynamic Testing of V/STOL Models." AGARD, AGARDograph 126, May 1968.

64. Cook, W. L., and Whittley, D.C.,"Comparison of Model and FlightTestData for anAugmented JetFlap STOL Research Aircraft." NASA TM X-62491, 1975.

65. Hickey, D. H., and Cook, W. L., "Correlation of Wind-Tunnel and Flight-Test Aerodynamic Data for Five V/STOL Aircraft." AGARD, Flight Mechanics Panel Meeting, Oct. 1965.

66. Vogler, R. D., "Ground Effects on Single- and Multiple-Jet VTOL Models at Transition speeds over Station- ary and Moving Ground Planes." NASA TN D-3213, Jan. 1966.

67. Campbell, J. P., Hassell, J. L., and Thomas, J. L., "Recent Research on Powered Lift STOL Ground Effects."

AIAA Paper 77-574, 1977.

68. Turner, T. R., "A Moving Belt Ground Plane for Wind Tunnel Ground Simulation and Results for Two Jet-Flap Configurations." NASA TN D-4228, 1967.

69. Butler, S. F. J., Moy, B. A., and Pound, T. N., "A Moving-Belt Rig for Ground Simulation in Low-Speed Wind Tunnels." Aeronautical Research Council, R & M., No. 3451, 1967.

70. Hackett, J. E., Boles, R. A., and Lilley, D. E., "Ground Effect for V/STOL Aircraft Configurations and Its Simulation in the Wind Tunnel." "Pt. I. Introduction and Theoretical Methods." "Pt. II. Experimental Studies." "Pt. III. The Tangentially Blown Ground as an Alternative to a Moving Ground, Application to the Ames 40 by 80." NASA CRs 114,495; 114,496; and 114,497, 1974.

71. Poisson-Quinton, Ph., and Christophe, J., "Special Ground Testing Facilities and Testing Techniques for STOL Aircraft." Von Karman Institute for Fluid Dynamics, Lecture Series 60, STOL Technology, Sept. 1973.

72. Hackett, J. E., Boles, R. A., Lilley, D. E., "Ground Simulation and Tunnel Blockage for a Jet-Flapped, Basic STOL Model Tested to Very High Lift Coefficients." NASA CR-137,857, 1976.

73. Crowder, J. P., Goldhammer, M. I, and Smyth, D. N., "STOL Aircraft Transient Ground Effects. Pt. II.

Experimental Techniques Feasibility Study." NASA CR-137,767, 1975.

74. Wolowicz, C. H., Bowman Jr., J. S., and Gilbert, W. P., "Similitude Requirement and Scaling Relationships as Applied to Model Testing." NASA TP-1435, 1979.

75. Gainer, T. G., and Hoffman, S., "Summary of Transformation Equations of Motion Used in Free-Flight and Wind Tunnel Data Reduction and Analysis." NASA SP-3070, 1972.

76. Paulson, John W., "Investigation of the Low-Speed Flight Characteristics of a 1/15 Scale Model of the Convair XV-58 Airplane." NACA RM-SL57K19.

77. Putnam, W. F., and Curtiss, H. C., Jr., "Low Speed Testing-Techniques for V/STOL Aircraft in the Princeton Dynamic Model Track." AIAA Paper 79-0334, 1979.

78. Putnam, W. F., "Tests on a Twin-Turbofan V/STOL Model in Ground Effect in the Dynamic Model Track_" Princeton University, Princeton, New Jersey, Final Report for Navy Contract N62269-79-C-0223, Jan. 1983.

79. Heyson, Harry H., "Wind Tunnel Testing of VTOL and STOL Aircraft." NASA TM-7850, 1978.

80. Maskell, E. C., "A Theory of the Blockage Effects on Bluff Bodies and Stalled Wings in a Closed Wind Tunnel." Royal Aeronautical Establishment R&M No. 3400, Nov. 1963.

81. Peitzman, F. W., "Determination of High Attitude Wall Corrections in a Low Speed Wind Tunnel." AIAA Paper 78-810, 1978.

82. Hackett, J. E., "Living with Solid Walled Wind Tunnels." AIAA Paper 82-0583, 1982.

83. Hackett, J. E., Wilsden, D. J., and Lilley, D. E., "Estimation of Tunnel Blockage from Wall Pressure Signatures: A Review and Data Correlation." NASA CR-152,241, 1979.

84. Snyder, L. D., and Erickson, L. L., "PAN-AIR Prediction of NASA Ames 12-Foot Pressure Wind Tunnel Inter- ference on a Fighter Configuration." AIAA Paper 84-0219, 1984.

85. Tyler, R. A., and Williamson, R. G., "Experience with the NRC lO-ft x 2O-ft V/STOL Propulsion Tunnel: Some Practical Aspects of V/STOL Engine Model Testing." NAE Quarterly Bulletin, No. 2, 1973, pp. 45-59, 61-75.

86. Hoad, D. R., and Gentry, G. L., Jr., "Longitudinal Aerodynamic Characteristics of a Low-Wing Lift Fan Transport Including Hover Characteristics in and out of Ground Effect." NASA TM X-34020, 1977.

87. "Conference on V/STOL and STOL Aircraft." NASA SP-116, 1966.

88. Carbonaro, M., "Interference Problems in V/STOL Testing at Low Speeds." AGARD-CP-174, No. 40, Oct. 1975.

89. Shindo, S., and Rae, W. H., Jr., "Low Speed Test Limit of V/STOL Models Located Vertically Off-Center."

J. Aircraft, Vol. 15, No. 4, April 1978, pp. 253, 254.

90. Shindo, S., and Rae, W. H., Jr., "Recent Research on V/STOL Test Limits at the University of Washington Aeronautical Laboratory." NASA CR-3237, 1980.

91. Steinle, F. W., and Pejack, E. R., "Toward an Improved Transonic Wind-Tunnel Wall Geometry- A Numerical Study." AIAA Paper 80-0442, 1980.

92. Sears, W. R., "Adaptable Wind Tunnel for Testing V/STOL Configurations at High Lift." J. Aircraft, Vol. 20, No. 11, Nov. 1983, pp. 968-974.

93. Christiansen, R. S., "A Large Scale Investigation of V/STOL Ground Effects." AIAA Paper 84-0336, 1984.

94. Henderson, C., and Walters, M., "Development and Validation of the V/STOL Aerodynamics and Stability and Control Manual." AIAA Paper 81-2611, Dec. 1981.

95. Kuhlman, J. M., and Ousterhout, D. S., "Experimental Investigation of Effect of Jet Decay Rate on Jet- Induced Pressures on a Flat Plate." NASA CR-2979, 1978.

96.

Kuhlman, J. M., and Warcup, R. W., "Jet Decay Rate Effects on Hover Jet-lnduced Loads." J. Aircraft, Vol. 17, No. 8, Aug. 1980, p. 605.

97.

Wulf, R., and Melzer, E., "Wind Tunnel Testing with Engine Simulation for V/STOL Airplanes," AGARD, V/STOL Propulsion Systems, No. 135, Paper No. 8, 1973.

98.

Margason, Richard J., and Gentry, Carl L., "Static Calibration of an Ejector Unit for Simulation of Jet Engines in Small-Scale Wind Tunnel Modes." NASA TN D-3867, Mar. 1967.

99.

Hoad, D. R., "Longitudinal Aerodynamic Characteristics of an Externally Blown Flap Powered Lift Model with Several Propulsive System Simulators." NASA TN D-7670, 1974.

100.

Nicoloff, C. B., and Weber, H. A., "Characteristics of an Ejector-Type Engine Simulator for STOL Model Testing." AIAA Paper 72-1038, 1972.

101.

Stewart, V. R., "Low Speed Wind Tunnel Tests of Ground Proximity and Deck Edge Effects on a Lift Cruise Fan V/STOL Configuration." NASA CR-152,247.

102.

Hunt, D., Clingen, J., Salemann, V., and Omar, E., "Wind Tunnel and Ground Static Tests of a 0.094 Scale Powered Model of a Modified T-39 Lift/Cruise Fan V/STOL Research Airplane." NASA CR-151,923, 1977.

103.

"Installation and Operation Manual, Model 1109, 12" Diameter, Tip-Turbine Fan." TM78-114, Tech Develop- ment, Inc., Dayton, Ohio, Oct. 1978.

104.

Eigenman, M. F., and Bailey, R. 0., "Development of the Propulsion Simulator- A Test Tool Applicable to V/STOL Configurations." Society of Automotive Engineers 770984, Nov. 1977.

105.

Bailey, R., Mraz, M., and Hiley, P., "The Design of a Wind Tunnel V/STOL Fighter Model Incorporating Turbine Powered Engine Simulators." AIAA Paper 81-2635, 1981.

106.

Harper, M., "The Propulsion Simulator Calibration Laboratory at Ames Research Center." AIAA Paper 82-0574, 1982.

107.

Smith, S. C., "Determining Compressor Inlet Airflow in the Compact Multimission Aircraft Propulsion Simulators in Wind Tunnel Applications." AIAA Paper 83-1231, 1983.

108.

Falarski, M. D., Whitten, P. D., and Harris, P. D., "Aerodynamic Characteristics of a Large Scale Model of a Highly Maneuverable Supersonic V/STOL Fighter: STOL Configuration." AIAA Paper 80-0234, 1980.

109.

Dawson, C. R., "Ground Effects and Control Effectiveness Tests of a 0.095 Scale Powered Model of a Modified T-39 Lift/Cruise Fan V/STOL Research Airplane." Boeing Document D 180-20391-1, Boeing Aircraft Company, Seattle, Wash., Feb. 1977.

110.

Juanareno, D. B., and Blumenthal, P. Z., "A Remote Millivolt Multiplexer and Amplifier Module for Wind Tunnel Data Acquisition." Pressure Systems Incorporated, Hampton, Va., NASA Lewis Contract, NAS3-22950.

111.

Cambra, J. M., and Tolari, G. P., "Real-Time Computer Data System for the 40- by 80-Foot Wind Tunnel Facility at Ames Research Center." NASA TN D-7970, 1975.

112.

Cambra, J. M., and Trover, W. F., "The Revolution in Data Gathering Systems." NASA TM X-62,452, 1975.

113.

Schoenster, J. A., Willis, C. M., Schroeder, J. C., and Mixson, John S., "Acoustic-Loads Research for Powered-Lift Configurations." NASA SP-406, May 1976, pp. 429-443.

114.

Strout, F. G., and Atencio, A., Jr., "Flight Effects on JT8D Engine Jet Noise as Measured in the NASA Ames 40- by 80-Foot Wind Tunnel." AIAA Paper 76-556, 1976.

115.

Soderman, P. T., "Instrumentation and Techniques for Acoustic Research in Wind Tunnels." IEEE Publica- tion 75 CHU 993 6 AES, 6th International Congress on Instrumentation in Aerospace Simulation Facilities, Ottawa, Canada, Sept. 22-24, 1975.

116.

Falarski, M. D., Koenig, D. G., and Soderman, P. T., "Aspects of Investigating STOL Noise Using Large- Scale Wind-Tunnel Models." Journal of the Canadian Aeronautics and Space Institute, Vol. 19, No. 2, Feb. 1973.

117.

Diedrich, J. H., and Luidens, R. W., "Measurement of Model Propulsion System Noise in a Low-Speed Wind Tunnel." AIAA Paper No. 76-91, 1976.

118.

Ahtye, W. F., "Wing and Flap Noise Measured by Near- and Far-Field Cross, Correlation Techniques."

AIAA Paper 79-0667, 1979.

119.

Atencio, A., Jr., and Soderman, P. T., "Comparison of Aircraft Noise Measured in Flight Test and in the NASA Ames 40- by 80-Foot Wind Tunnel." AIAA Paper 73-104.7, 1973.

120.

Cook, W.L., and Hickey, D.H.,"Correlation of Low Speed Wind Tunnel and FlightTestData for V/STOL Aircraft." NASA TM X-62,423, 1975.

121.

Falarski, M.D.,and Moore, M.T., "Acoustic Characteristics of Two Hybrid Inletsof Forward Speed," AIAA Paper 79-0678, 1979.

122.

Moore, M.T., "Forward Velocity Effects onFan Noise and theSuppression Characteristics ofAdvanced InletsasMeasured in theNASA Ames 40-by80-Foot Wind Tunnel." NASA CR-152328, 1979.

123.

Margason, R.J., "The Path of a JetDirected at Large Angles to a Subsonic Free Stream." NASA TN D-4919, 1968.

124.

Werle, H.,"Flow Visualization Techniques for theStudy of High Incidence Aerodynamics." AGARD VKI Lecture Series 121, 1982.

125.

Merzkirch, W.,ed.,"Flow Visualization Proceedings of theSecond International Symposium onFlow Visual- ization."Crowder, J. P., "Fluorescent Minitufts of Nonuniform Surfaces Flow Visualization." Proceedings of theSecond International Symposium onFlow Visualization, Bochum, West Germany, pp.663-667, Sept. 1980.

126.

Imaichi, K., and Ohmi, K., "Numerical Processing of Flow-Visualization Pictures - Measurement of Two- Dimensional Vortex Flow." J. Fluid Mech., Vol. 129, Apr. 1983, pp. 283-311.

127.

Hunt, J. C. R., Abell, C. J., Peterka, J. A., and Woo, H., "Kinematical Studies of the Flows around Free or Surface-Mounted Obstacles: Applying Topology to Flow Visualization." J. Fluid Mech., Vol. 86, May 1978, pp. 179-200.

128.

Kao, H. C., and Burstadt, P. L., "Flow Visualization and Interpretation of Visualization Data for Deflected Thrust V/STOL Nozzles." NASA TM-83554, 1984.

129.

Mueller, T. J., "Smoke Visualization in Wind Tunnels." Astronaut. Aeronaut., Jan. 1983, pp. 50-62.

130.

Naeseth, Roger L., and Hoad, Danny R., "Upwash Angles Near Engine Inlets of an Externally Blown Flap STOL Transport." NASA TN D-8091, 1975.

131.

Everett, K. N., Gerner, A. A., and Durston, D. A., "Theory and Calibration of Non-Nulling Seven Hole Cone Probes for Use in Complex Flow Measurement." AIAA Paper 82-0232, 1982.

132.

Orloff, K. L., Snyder, P. K., and Reinath, M. S., "Laser Velocimetry in the Low Speed Wind Tunnels at Ames Research Center." NASA TM-85885, 1984.

133.

Orloff, K. L., and Snyder, P. K., "Laser Doppler Anemometer Measurements Using Nonorthogonal Velocity Components: Error Estimates." Appl. Opt., Vol. 21, Jan. 1982, pp. 339.

134.

Beranl, L. D., and Sarohia, V., "Entrainment and Mixing in Thrust Augmenting Ejectors." AIAA Paper 83-0172, 1983.

135.

Reinath, M. S., Orloff, K. L., and Snyder, P. K., "A Laser Velocimeter System for the Ames 40- by 80-Foot and 80- by 120-Foot Wind Tunnels." AIAA Paper 84-0414, Jan. 1984 (also NASA TM-84393, 1984).

TABLE 1.- EVALUATION OF FIXED FRAME FACILITIES (REPRINTED FROM REF. 10) Facility Objectives NASA/ NASA/ NASA/Ames NASA/Ames GE open throat requirements Langley Lewis 7' x 10' 40' x 80' anechoic wind tunnel Nozzle 25 Ibm/sec 25 Ibm/sec weight flow 25 Ibm/sec 17 Ibm/sec 25 Ibm/sec with modi- with modi- 42 Ibm/sec fication fication Nozzle size 4" - 6" ? 4" - 6" 4" 6" 6" Exhaust gas Burner Ambient - total tempera- 520°R-2500°R needs to be Ambient Ambient 1600OR Ambient - 3000°R ture range provided Nozzle pressure ratio 1.5 - 4.0 ? 1.5 - 4.0 1.0 - 2.5 1.0 - 2.5 1.5 - 4.0 Free stream 0-400 ft/sec 0-400 ft/sec 0-220 ft/sec 0-400 0-325 velocity range ft/sec ft/sec 0-330 ft/sec Facility Anechoic Anechoic Hardwall Scottfelt Partially wall type treated Anechoic Proven nozzle Extensive Will have Extensive Extensive Feasibility Performance performance modifi- capability modifi- modifi- measurements system + 1/2% cations by Dec. 1974 cations cations to be studied accuracy Facility 1974, 1975, 1974, 1975 1976 1974, 1975, 9 weeks/yr 1975, 1976 availability 1976 1976 1976 Duration of operation Continuous Blowdown Continuous Continuous Continuous Continuous Capable of per- Would Would Designed to forming laser Large window Minor require Yes require accept laser velocimeter high quality modifi- modifi- modifi- measurements optical glass cations cations cations velocimeter TABLE 2.- REDUCED VELOCITY SCALE (FROM REF. 34) Vo/V j 0.05 0.075 0.10 0.15 0.20 0.30 0.50 C (S/Aj = 40) 20 9 5 2 1.25 0.6 0.2 Vo Full 20 30 40 60 80 120 200 scale, 400 Vj m/sec Vo Model 20 scale, 400 267 200 133 I00 67 40 Vj m/sec --- 1.0 0.67 0.5 0.33 0.25 0.17 0.1 Vo model scale Vo full scale/ TABLE 3.- ACCEPTABLE LIMITS TO WALL CORRECTIONS (FROM REF. 88) Maximum acceptable Moderate No corrections Parameter corrections corrections As ± 5 ° ± 5 ° ± 1/2 ° 1 ± 5% qc/q I ± 10% I ± 10% ± 1/2 ° Ai t ± 5 ° ± 2 ° 1 ± 5% qt/qc I ± 10% 1 ± 5% ± 1/2 ° Aiw ± 2 ° ± 1/2 ° ± l°/semispan d(Aiw)/d(y/b) ± 5°/semispan ± l°/semispan ± 5% Aq/qc ± 10% ± 5% CORRESPONDING LIMITS FOR A JET FLAPPED WING OF ASPECT RATIO 6; B/H = 4/3 Moderate Maximum corrections corrections Normal mounting at the center of a 4:3 test section b/B = 1/2 CL < 6.3 CL = 14.1 b/B = 3/4 CL < 0.81 CL < 4.14 Semispan mounting on the floor of a 4:3 test section (b/2)/H = 1/3 CL < 15 CL < 20.4 (b/2)/H = 1/2 CL < 4.86 CL < Ii.0 (b/2)/H = 2/3 CL < 2.4 CL < 5,82 TABLE 4.- JET THRUST SIMULATION (FROM REF. 5 WITH ADDITIONS Other Air supply Principal simulation Comment Nacelle type requirement simulations capabilities Flow None .Inlet geometry .Dual flow •Simple • Inlet Vi/V_ at .Good for aerodynamic one Mach No, drag studies Blown jet Large ,Exhaust nozzle ,Dual flow •Simple operation geometry .Hot gas .Erroneous inlet • Gross thrust contribution Ejector Moderate •Inlet geometry/or ,Dual flow .Inlet and exhaust • Exhaust nozzle flow not simulated geometry/and simultaneously • Gross thrust Turbine Small ,Inlet geometry ,Sensitive mechanism powered or exhaust ,Difficult to nozzle geometry simulate inlet and • Inlet Vi/V and exhaust flow not gross thrust simultaneously • Gross thrust and not inlet Vi/V Gas turbine Small -Inlet and exhaust .Dual flow .For subscale use powered nozzle geometry bypass ratio may • Inlet Vi/V plus not always be gross thrust simulated • Exhaust Mach No. .For subscale, engine and temperature is often over size

TABLE 5.- LEVELS OF SOPHISTICATION: LARGE-SCALE MODELS

Generally Increasing Model Cost >

Factor I 2 3 4

Airframe

Approximated Scaled Full-scale

configuration configuration configuration

Aircraft

Propulsion Simulated Engine used Including

propulsion butsmall some structure

Controls

Scaled Same engine

Dynamics

Electronics

Other

Examples

Model type 1 2 3 4

Complete High_ (0.4scale) QSRA model AV-8B XFV 15

configuration (VATOL models) 40by80 40by80

Q-fan simulation Basic USB/EBF G698 AV-8B

of tilt nacelle research models 40by80 (static)

M-260 XFV 12A

97powered Tether and

ejector model static

Semispan Aug Wing Cooling drag Cooling drag NA

40by80

Components Thrust reverser Acoustic test 698 nacelle NA

40by80 "D"nozzle

= 50 °, FLAPS DOWN, GEAR DOWN, TAIL OFF -TOTAL MODEL T._ __ MODEL FORCE /"1" \ MOMENT Cm L TOTAL DIRECT DIRECT THRUST THRUST THRUST Cj COMPONENT COMPONENT C L MODEL MOMENT LESS THRUST MODEL LIFT LESS AND RAM THRUST COMPONENT Cm A

7-°/

I I I 0 i -2 -1 0 1 2 0 10 20 30 0 -1 e,deg CD Cm Figure 1. Induced effects for transition flight.

1.6 AV-8B AV-8B -- 2.1 m (7 ft) diam. MODEL (FULL SCALE) FULL SCALE _ _,_-'_'-_- / 15% SCALE 1.2 .8 60 I FULLSCALE ^!,_ _ I 6755 Ib CL ,'_ 15% SCALE I _-4g-_ 4g'-E]-'E] .4 20 -- 1/5 SCALE i

i I I I I I

i i i i i i [ i -,4 .2 .4 .6 .8 1.0 0 4 8 12 16 20 24 28 -4 I/T c _x, dpg Figure 3. Effect of scale on duct inlet stall Figure 2. Comparison of large- and small-scale test data for the AV-8B (Ref. 8). boundary (Ref. 4).

FULL o } 1/12 SCALE o 0 : FULL} FULL SCALE 3.6 - 1.59 m (5.2 ft) diam. FAN, 17.8 cm (7 in.) CHORD, 3.2 .;_ 1.0 _//R = 1.4 x 106 .8 2.4 C L 2.0 .6 /3 v,deg . . " . " . FAN, .4 _ O_ -_ 20.265 :m(_1.83;;)i1._ cHdiRD, 20 R = 0.22 x 106 1.6 _ SLAT NOSE ROTATION .2 Z_ 30 1.2 I_ 40 I I I I I .8 I I I I I I I I 0 .2 .4 .6 .8 1.0 4 8 12 16 20 24 28 32 Fx/LI3 v = 0 a, deg Figure 5. Effect of scale on the lift character- Figure 4. Performance of large- and small-scale exit vanes (Ref. 4). istics of the FIIlA airplane landing configuration (Ref. 4).

b = 1.91 (6.25) MOUNTED IN 40 BY 80 INSERT - LANGLEY V/STOL TUNNEL H Figure 9. Rotor test rig mounted in the Ames 40 by 80 tunnel.

A M AL b = 11.62 (38.18) MOUNTED A T A T IN 40 BY 80 1.0.5 _-E_:_WING MOMENTUM AREA RATIO O • XV 3 -"_'\\\'"_ ..... O

__ .2 - ",,_ o '_",_,_, D II vz3

Figure 6. Externally blown flap models with the same wing-flap geometry.

= ,_ 0 • xc 142

.05 =-- • _ LIFTING ELEMENT a • XV4A " _lk_ AREARATIO I_ I_ XVSA k- .02 P_ • AV8B SCALE .01-- - =_' AM/AT a • L/C3 SMALL b : 1.91 (6.25)

<_005 --= " " ='", o • usB4

_ LARGE b = 11.62 (38.18) .002 I'i; 1Or" .001 i, Jh.,l ,, lJinil r| ,h.i_d lind i E] • G698 102 103 104 105 106( DISK .6- , N/m2 / LOADING lllhllli l llhllll l lillml U llJlilll I c L / 2 10 100 1000 10000 .4-- / / Ib/ft 2 ,2-- (a) Lift and momentum area sizing.

I I I I 1 I o 16 32 -16 0 -.8-.16 2.4 0 XV3 c m I-I vz3 I I I I -,8 0 .8 1.6 1.o_-;, Q ,, _ XC 142 CD • ;_' " '_'"_',,,_,_,,,_,Q(_ .......... Z_I%. xvXV4AsA Figure 7. Comparison of wind-tunnel test results m .1_-- P_ AV8B on the externally blown flap models of Fig. 6 .a.05_. Q L/C3 (Ref. 4).

.02_ Q USB 4 01/I ilh.il i i,l..l .,,I..I i,ilillil i iD G698 102 103 104 105 106| DISK N/m2 _1" LOADING l llhi.l l lilmll i iihlill lllhilll i 2 10 100 1000 10000 Ib/ft 2 (b) Wing span sizing.

Figure i0. Sizing parameters (Ref. 4).

Figure 8. Large-scale J97-powered model of the GD205 two-engine fighter configuration.

3O Figure 12. The Air Force ejector mounted in a Figure 11. Bell/NASA ejector flap wing semispan model mounted in the Ames 40 by 80 wind tunnel semispan installation in the Ames 7 by 10 tunnel: (Ref. 12).

6f = 90 o (Ref. 13).

LARGE SCALE • AERODYNAMIC PERFORMANCE • STABILITY AND • NOISE _ FARFIELD CONTROL _, NEARFIELD EXHAUST _ _ GAS _ INGESTION DEF LE CTOA_ Y_ _ T_E A D_JA _ D SURVEYS FLUCTUATING • PRESSURE DISTRIBUTION LOADS • GROUND EFFECT SMALL SCALE COMPONENT TESTS • INLET • INGESTION • DEFLECTOR PERFORMANCE • LOADS COMPLETE MODEL TESTS • PERFORMANCE • STABILITY AND CONTROL • PRESSURE DISTRIBUTION • INGESTION Figure 13. Measurements using large- and small- scale models (Ref. 4), USB RESEARCH COST $ 800K 2 W.T. TESTS I STATIC TEST 40 BY 80 ENGINE AIRFRAME SUPPORT J _ CONTRACTOR • STATIC TESTS • MODEL DESIGN • WIND TUNNEL TESTS • CONSTRUCTION • SUPPORT • INSTRUMENTATION +< .>:.: An in-house basic research project.

(a)

VTOL RESEARCH COST $1500K 2 W.T. TESTS 1 STATIC TEST (LIFT-CRUISE FAN)

I

' 'A I

NASA 40 BY 80 AIRFRAME

: °'_;2_'_.NS CONTRACTOR

• TEST MANAGEMENT • STATIC TEST • DESIGN • MODEL CONSTRUCTION FACILITY • TEST SUPPORT • INSTRUMENTATION • WIND TUNNEL • DATA ANALYSIS • SUPPORT • CONSULTATION (b) A joint service R. & D. project.

1 WoT.

COST > $ 3000K 1 STATIC TEST HARRIER • OPERATIONS = CONTRACTOR NASA 40 BY 80 I I AIRFRAME • RESEARCH

J I

• TEST • STATIC TEST • MODEL • TEST • TEST MANAGEMENT MANAGEMENT FACILITY • WIND TUNNEL SUPPORT • SUPPORT (c) A military development program.

Figure 14. Organization and costs of some large-scale V/STOL wind-tunnel investigations (Ref. 4).

FAN SECTION I" C 129 m / B,I , _ B D I

i ---

VA.ES -IG_-_--_ ..... _:_fr

j .... £i "

I Ilfifo

I I__ +==SEOT'°NS-'> _i,m T"ROTT'E

- , C, d : .;;i;ii AIR EXCHANGE _B _-"E HEAT EXCHANGER HATCHES _..A__ ....... ---_-,-- --.-X.._: i .... _r ..... BALANCE "_"<_<_ Illt ............

oo.

E--E C-C D-D B-B A-A WORKING SECTION CROSS-SECTIONAL SHAPE & SIZE, _%_! _ , , .

mXm l_i'tiiii_li 6 611117 6 it_11<1_ III 9.5 8 6 8 TYPE CLOSED CLOSED CLOSED OPEN USABLE LENGTH, m 15 16 9 20 62 117 153 80 MAX. AIR SPEED, m/sec CONTRACTION RATIO 4.8 9.0 12.0 9.0 MAX. MODEL SPAN, m 6.5 --_ 5.5 _ 4.2 _ 4.0 Figure 15. DNW wind tunnel (Refs. 23, 24).

I-_1_:' _' !']"i:'i']"_ :"":"_;i I''I 'i i, ......... AIR INo.T_ ,_l i

_F E:::_.,,ou.,,:.;_f,,c.,o. ,_..,,NI 11

SCALE m=s;====_:===s_=_ ENGINE TEST BENCH ill (a) General layout.

mF---___

\ f

am

-- i_ _i

I PAROI //I I_11 m No.2 i// I I _ II (b) One of the test sections.

Figure 16. ONERA-Modane 8M Facility (Ref. 26).

AIR EXHAUST _ ......... THIRDDIFEUSER SECOND FLOW- CONTROL_ FIRSTIDIFFUSER

_'"_c_c°_s_s_c_'°"li i _,.,<ow-

CHA_BI IC._BER

Figure 17. Planview of the Langley 4 by 7 tunnel (Ref. 27).

PLENUM AIR EXHAUST TOWER LOW SPEED DIFFUSER TEST SECTION BELLMOUTH AIR INLET TOWER FAN SECTION HIGH SPEED DIFFUSER TEST CELL MODEL SHOP COMPUTER ROOM AIR COMPRESSOR BLDG.

COMPUTER GRAPHICS ENGINEERING OFFICES MODEL ASSEMBLY AREA Figure 18. Boeing V/STOL wind tunnel (Ref. 28).

12x24m (40 x 80 ft) TEST SECTION 100 m/sec -* 150 m/sec (200 knots _ 300 knots)

x

(80 x 120 ft) TEST SECTION

50 m/sec //_J/",\ _I:L_'I "% ill

(100 knots) _/_2_/____11_ _11 27 MW -* 100 MW (36,000 hp -* 135,000 hp) DRIVE Figure 20. Ames large-scale wind-tunnel system Figure 19. The test section of Boeing V/STOL wind tunnel: open-throat mode. (Ref. 29).

ELEVATION --90 ft 3 in.

--66 ft 3 in.

-- 49 ft 3 in.

"HIGH BAY"

" 11 . . AC_ES_ _ONTROL, I III

II

--26 ft 6 in, FLOOR LEVEL 2 ft Figure 21. Elevation of Ames 40 by 80 test section.

H Figure 23. Sketch of Ames 80 by 120 test Figure 22. Complete model (G698) installed in the section, Ames 40 by 80 test section.

(a) Compressor noise studies:hybrid inlet. (b) 0.11-scale G698 hover tests.

Figure 24. Special testsetups in theAmes 40by80tunnel.

° ....

434.5 _"1 ALL DIMENSIONS IN FEET Figure 26. Langley 30 by 60 test section.

Figure 25. Langley 30 by 60 wind tunnel.

• ....... i

:i_,_,_., ......

iiiiiiiiiiiiiiiiiiiiii

:::::::::::::::::::: ::::::::::::::.y ii!iiii!_}}!iT ::::U:::::: Z_z_ZzZz _ .:,:.:,:.:.:.:,:.:.:,:.:.:.:-:+::c. :: i ::;;_, _ i.:+: : ::.

;;::._;;;;;:;;;:;:;;::;;;;:::.:," ::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::: _i#_ :_"_;LLiiiiiiiii_i__i!!ii_!_ii_i!i!i_i!_!_!_!i_i_i!i:i:i_i_i:i_i:i:i_:_:i:i:i:i:i:i:_:_:i:i_i:i:i:i_:i:i: :_::::: Figure 27. The test section of the NRC 9-m wind tunnel.

DRY AIR STORAGE

!;:-STAGEAX,A<,<OW TRA I R"E"

I AUX'<,AR, EOU,PMENT

(_,,._/f'_k"_ /_ '_... MOTORS &_:_'_r_li, _ I__'_ 8"BY 7-FOOT SUPERSONIC _______EST SECTION wE.

11-BY TEST11"FOOTsEcTIoNT_ANSONIC _ VAC_

. ,_.R >_r r "%,_... ,T,(,,%o

,1-STAGE AXIAL/" _ FLowANv2'2SI,2_ ECTOR '":_w%,t x'A'_" VA,.

9-BY 7-FOOT SUPERSONIC TEST SECTION Figure 28. Ames Unitary Wind Tunnel System.

Figure 29. BAC 5.5-m (18 ft) open-throat wind Figure 30. General arrangement of circular wing tunnel (Ref. 34). and ground plane (Ref. 35).

2.8 0 0.45 I-I 0.95 2.4 _ 1.88 Z& 3.76 3.2 II C'u 2.0 V 5.42 ----- LARGE < 1.2 1.0 ,8 1,6 _ Z i | i i

0 .2 .4 ._ ._ ,.0 1.2 _.4 _.6 _._ 2.° _.2

h/D o Figure 31. Effect of C. on the circular wing . . ._ • . 0 0 vertlcal thrust vamatlon with altltude: _ = (Ref. 35).

(a) 3/4frontview. (b) Rear view.

Figure 32. Qfanand a pitch-control vane mounted in theAmes 40by80tunnel (Ref. 37).

+> (b) Small-scale tests.

(c) Semispan tests.

(d) Dynamometer tests.

(e) Full-scale wind-tunnel tests.

(f) XV15 in flight.

Figure 33. XV-15 development.

4O AC L -2 [ -4 z$CD -2 I I I I I I -4 Figure 36. Jet-in-cross-flow body mounted in Ames 7 by 10 wind tunnel (Ref. 46).

8 L (_, deg

y-._',_ o 45

............. iiii_i_i_,,:_i:_,,i :,:_i:_i:_i_i_i___:'ii:_ '_. _ ................... _,_

4 I- u._-_ I-I 60 _C m -4 -8 I I I I I I -10-30 -20 - ,0 0 10 20 30 _v, deg Figure 34. Effectiveness of the pitch-control vane Figure 37. Jet-in-cross-flow installation in the Langley V/STOL wind tunnel (Ref. 47).

for the subsonic tilt nacelle aircraft: Cj = 10 (Ref. 37).

Figure 38. Installation of high-lift wing model Figure 35. Above the wing inlet study in the Ames and turbine fan between two-dimensional blowing 11-ft tunnel (Ref. 42).

walls (Ref. 49).

,,, ,,,,, ,,,,,, %,,%,,,,,,%,, ,,,,,,,,,, ¸,¸,¸,,,¸¸¸,¸,¸,¸,¸,¸,% ,, %% H (a) Wind-tunnel installation. (b) Details of model.

Figure 39. A quasi-two-dimensional ejector flap model (Ref. 50).

_WIN_ NO_.7_L _ = TUNN_ L _t.0_R %,,,,¸ R _TA_T I I_N : ............ :i!!ii!i i! ¸¸ • . L_ ....

Figure 40. The cruise blowing model in the Boeing Figure 41. Semispan mount for a wing-fuselage-tail transonic wind tunnel (Ref. 52). configuration (Ref. 53).

I

.>:.:.:+:.:.>:.:.:." : FUSELAGE SUPPORT EXTENSION

_PEDESTAL EXTENSION

I REMOVED

I

(a) Mounting details. (b) View of installation in the Boeing Vertol wind tunnel.

Figure 42. Wing-fuselage semispan mount with the fuselage nonmetric (Ref. 54).

::!

Figure 43. Twin-engine, high-speed model mounted Figure 44. Sting mount of an above-the-wing inlet in the Ames 12-ft tunnel (Ref. 56). model in the Ames 11-ft tunnel (Ref. 43).

Figure 47. Twin-tail support for a generic fighter

model powered bydirectblowing (Ref. 59).

Figure 45. CMAPS-powered model in theAmes 11-ft

transonic tunnel (Ref.57).

_MODI !L BASELINE (NO ADAPTER) _ SHORT STING .__

\__

_LONG STING _Cp = CpWlT H - CpBASELIN E

.20

=J ADAPTER (NO ADAPTER)_ mr .16

MACHO., [

z m ,12 MODEL TE _ / ,,_ .08 AFT METRIC BREAK _,_ j_ mr Figure 48. A sting mount installation in the Ames .04 40 by 80 for high-angle-of-attack studies (Refs.

ro- 60, 61).

W 0 LONG STING

l , I

-,04

1'0 2'0 _'0 4'0 50 _0 70 8'0 9_ 1_01_0

MODEL STATION, in.

Figure 46. Sting-length effect on adapter-induced pressure increments for the CMAPS-powered model of Fig. 45 (Ref. 58).

(a) Mounted in Ames-Army Air Mobility Command 7 by (b) The H-126 aircraftmounted in theAmes 40 i0 I/7-scale (Ref. 62). by80.

Figure 49. Wind-tunnel mounting of theH-126.

Figure 50. Composite model rig forjet-nacelle

model withwing-fuselage inverted (Ref.63).

COUPLING (STRUT) _UPPER FAIRING YAW(CHAIN) DRIVE

ltllll ) POWER POD FAIRING

YAW MOTOR AND GEARBOX J AIR MOTORS COMMON POWER- BOX (GAUGES) POWER SL SCAVENGE ANGLE CIOU SLIP-RING AIR EXHAUST ADAPTOR) Figure 51. Pedestal-mounted model on the Vertol power pod (Ref. 28).

Figure 52. Aircraft installation in the Langley 30 by 60.

(b) 3/4 rear.

(a) 3/4 front.

Figure 53. Full-scale AV-8B model mounted in She Ames 40 by 80 (Ref. 19).

TURNERS CRITERIA: MOVING BELT ADVISED FOR FULL SPAN HIGH LIFT DEVICES, REF. 68 / I0

I / I / / I I / //////Jz_ ___

MOVING BELT _ _ - _ _" p- Z

MOVING BELT

MJ - 6 C) u.

_?_ BLOW?NRG JET _ o o 4 _J _ _ \ _ ,ACCEPTABLE / .4 .8 1.2 1.6 2.0 2.4 TRAILING-EDGE HEIGHT/WING CHORD Figure 54. Criteria for ground-effect simulation (Refs. 29, 68).

LEADING-EDGE DETAILS UNNEL C L

) \\ _1o ._

| \\ +_-----_-_----7/j ' 24! cm

I1_---187 cm---_ I

_1, _+) .r- .... "-:r------ (+)1 H

co/. .= --j

w.ooE.A,..°o

"] I [-"-'_ 'ANGLE-OF-ATTACK L_ MECHANISM SIDESLIP SPINDLE Figure 55. Turner's moving-belt installation (Ref. 68).

VB/V o /_. RE-ENTRY

o o

[] 0.26 Z_ 1.04 E '-' 6 .:-

f

l

J

_DRIVE ir'_ H/_ '_ \ STING AND MODEL ,k_ 0 .6 1.0 1,2 .2 .4 .8 ,_ _-/ MOVII_G BELT V/V 0 BOUNDARY LAYER SCOOP Figure 56. Effect of belt velocity on boundary- Figure 57. Moving-belt installation in the Boeing layer profiles: station 187; see Fig. 55 (Ref. 68). V/STOL wind tunnel (Ref. 28).

VIEW A- A

TYPICAL MODEL INSTALLATION

/

-- --I -_- _---} /¢ I I I I_V-l-4_l I I V _ ,

__ _____ _, \ _L_,-,_--T ..... _<-_I

.... i

¢ TRAVEL ) | LIMITS BALANCE ROOM BELOW FLOOR Figure 58. Ames 40 by 80 proposed floor BLC design (Ref. 70).

_PPORT MOVING BELT _PORT

-I II'_H I

BLC ON FLOOR T NEGATIVE TAIL STALL A iX [3 MOVING BELT AT HIGHER ANGLES • • • BLC FLOOR TAIL STRIKES GROUND AT HIGHER ANGLES I I_l F//////////////A C# = 3,0 _/J/J/-/_ Cp = 1.0 TAIL ON TAIL ON TAIL ON _/// j'L ,

!

/ /// I //

(_.... rW,NG

, gTw'NG

c_ 8

._ _IWAKE

AI _WAKE _.',_,

J

, __\N _ , 0 - -2 -3 -1 -2 -3 -1 -2 -3 C m Cm Cm Figure 59. A sampling of data comparing the use of BLC on the wind-tunnel floor with that of a moving belt: h/c = 1.0 (Ref. 72).

FLOW TUNNEL FLOOR

/

mm PITCH SERVO VALVE SLIDE I VERTICAL POSITION REVISE EXISTING CALIB. SCALE BALCONY VER" SUMP PUMP ACCUM / / S I ANCHOR 8' x 11' CONCRETE FLOOR, 2 _ LEVEL Figure 60. Transient ground-effect support design for the Ames 40 by 80 (Ref. 73).

5O

m _ _ _ iii _ CONTROL CABLE

'

/ .J/m_- ' .

Y CABLE ItOLLAND _ II j_ _ t_'" _ OPERATOR II II YAW PILOT (a) Test setup in test section.

DIGITAL FLIGHT CONTROL POWERED DYNAMICALLY- PILOT CONTROLS COMPUTER AND RECORDERS SCALED MODEL • ELECTROPNEUMATIC • VISUAL REFERENCE • PROGRAMMABLE CONTROL ACTUATORS LAWS • PENCIL-TYPE CONTROL • VARIABLE GAINS AND • ONBOARD DATA SENSORS Wl3"H TRIMMER GEARING - MOTIONS • SELECTABLE CONTROL - CONTROL POSITIONS MODES • SELECTABLE FEEDBACKS - PROPORTIONAL • FAILURE MONITORING - FLICKER (b) Description of system features.

Figure 61. Free-flight operation in the Langley 30 by 60 wind tunnel (Ref. 74).

h (IGE), DATA SOURCE it h (OGE), cm (in.) cm (in.)

O DMT (PRINCETON) 20° 42 (16.5) 9.6 (3.8) r-I 40x80 TUNNEL (AMES) 20° 42 (16.6) 9.6 (3.8) Z_ 7x10 TUNNEL (LANGLEY) OFF 22.6 (8.9) 9.6 (3.8) _" 7x10 TUNNEL (15x17 SECT. OFF 22.6 (8.9) 9.6 (3_8) WITH MOVING BELT) 2.0 -6j =60°; _f=40°; CT=0.90 OGE IGE 1.6 .J

I

z 1.2

I

,,I

I

,,=

Figure 62. ,An X29A model in flight in the Langley 30 by 60 tunnel (Ref. 75).

I- .8

I

LL

I

I

I

/ NOTE: ALL GROUND CLEAR-

ANCE ALTITUDES ARE 1

CONVERTED TO EQUIV- ALENT 0.1 MODEL SCALE -1.2 -,8 -.4 DRAG COEFFICIENT, CD Figure 64. Comparison in ground-effect of results using the Princeton Track with that from three other facilities: tilt-wing model; h measured using Princeton model scale (Ref, 77).

(a) Model support.

(b) Grumman 698 model at wheel height.

Figure 63. Use of the Princeton track for ground- effect tests (Ref. 78).

(a) PAN AIR model of Ames 12-fttunnel (frontview).

--_ L TEST _1

--_ P- SECTION-"] v o V0 --- \ :

f

--_" UPSTREAM BUMP WING WAKE DOWNST R EAM TUNNEL SURVEY SURVEY WAKES PLANE PLANE

(b)

Schematic of complete PAN AIR model; section shown at plane of symmetry.

Figure 65. Evaluation wall constraints using potential-flow paneling method (Ref. 84).

ASSUME De/h = 0.143

Ve>0.187 Ve=0.187 Va=0.121 Ve=O.086

FLOOR STAGNATION POINT Ve = 1.31 (De/h)= 0.187 MINIMUM SPEED TEST LIMIT Vemin = 0.65 • (Ve) = 0.121 Ve< Vemi n /_ "_- VORTEX FORMATION Figure 66. Minimum-speed testing limit for a configuration with two side-by-side lifting jets at h/D e = 7 (Ref. 5).

_7 MAX. ACCEPT. CORRECTION _' MODERATE CORRECTION 10.0 • NO CORRECTION O CONVENTIONAL LIMIT A%= 2 -- A=0 o -- -- - A = 45 °

\

\

CL/A

\

1.0

\

\

.1 I I I i i I i I 0 1/6 1/4 2/6 3/6 4/6 3/4 5/6 1 b/B Figure 67. STOL testing limits: Di/L - _! model in center of wind tunnel; B/H = 4/3 (Ref. 88 U, SIDE WALL U, ROOF-FLOOR .12 .6 MEASURED MEASURED O CALCULATED O CALCULATED .08 .4 .04

o_

<_

i -,2 -.04: i I I I I I I I I i I -.8 -.4 0 .4 .8 1.2 1.6 -.8 -.4 0 .4 .8 1 2 1.6 x/b x/b (a) Wall pressure signatures.

THEORETICAL FLOW MODELS MEASURED TUNNEL PRESSURES

I

I 1 I I

EFFECT OF P EFFECT OF F I I ON SIDEWALLS ON ROOF/FLOOR SIGNATURES

IROOFANOFLOORI S,GNATURES I I SIDEWALL ]

I I

®1 DETERMINE r I

I REMOVE EFFECT OF F =_

ON SIDEWALLS iOUTPUT

r

I

EFFECT OF Q ON SIDEWALLS I FROM KNOWN P I

, IEVALUATE_ ]'

I ; BLOCKAGE I

(!)! DETERM'NEO

I I i EVALUATE I I I FROM KNOWN Q I .... .J L= ....

(b) Block diagram of procedure.

Figure 68. Evaluating wall corrections using wind-tunnel surface pressures (Ref. 82).

Vo .7'/1 /

Figure 69. Adaptable wall concept for V/STOL wind- tunnel testing (Ref. 92).

LARGE SCALE DATA

_L d12'25

T =-0"013(_) -2"3 (_10.05 (I_D I D/d = 7.93 -.2 k- <3 -.4

//

/

a) I I I I I -.6 30 I-10 SQUARE FLAT PLATE DRAG DATA LARGE SCALE DATA ------ NPR = 1.40 _.m NPR = 2.65 _L _o.o15r h/d ]-[2"2-0'24(NPR-1)]

-f=

D/d = 7.93 2520 (2 SOURCES)

T =_1o -,2

<3

,// /

-.4 0 I : I I I I -5 I I I I I I I b) I I J I I -.6 0 1 2 3 4 5 6 7 S/C, percent LARGE SCALE DATA (a) Wake blockage correction to flat-plate drag ------ NPR = 1.40 _--- NPR = 2.65 data assuming CDc = 1.17.

[_1-1"59

A L = -(0.00125Did + 0.0185)Csi --_C) V/STOL FIGHTER TEST METHOD T

LD dJ

---- -O MASKEL METHOD Csi = 1.173 - 0.2495 In(NPR), NPR < 2.0 = 1.061 - 0.0889 In(NPR), NPR >/2.0 10- D/d = 7.93 d n/ • 6 - -.2 i- #4.

<3 2 c_-c- -.4

oil-

-10 0 20 40 60 80 100

F/

_,deg (c) (b) Comparison of Maskell wake blockage correction -.6 I I I I with method developed for V/STOL fighter model test 2 3 4 5 applied to test run. h/d Figure 70. Blockage corrections for high-angle-of- Figure 71. Comparison of large-scale data with attack model (Ref. 60).

predicted lift losses (Ref. 93).

17-;-'/- ..-_ Z_

Anl/ ,_,-" J

I0 1.0 ,8 z/D e //

DF_o °

0 NO PLUG FLAT PLUG

"_\ [] DOWN 1.375 D n /

[] DOWN 1.375 D n .4 A DOWN 0.875 D n I_ DOWN 0.375 D n FLUSH .2 I I I I 0 2 4 6 8 10 12 14 3 6 9 12 s/De x/D e (a) Dynamic pressure decay. (b) Jet centerline trajectories.

Figure 72. Effect of a flat plug on characteristics of a round jet: R = 6 (Ref. 95).

Engine £imu/atJon ] Figure 73. Classification of principal engine simulation techniques (from Ref. 97).

(a) General model arrangement. (b) Ducting and nozzle details.

Figure 74. Hawker P.1127, 1/lOth-scale model (Ref. 63).

CONICAL SCREEN

/

UP _//t NOZZLE

PLENUM EXIT PROBE ______EXTENSION (I.6D) CROSS SECTION c::mI.//_q-_.E _//_ o o o c::_ | I_

o:ii:oiooo:ii

o ° o o

TURBULENCE _ _ TEMPERATURE

_ POWER

SCREENS _ .... ]

_BLOCK

I______ _ _ ____2:1] _PE R FORATED

_PLATE t AFT NOZZLE PLENUM EXIT AIR SUPPLY LINE OF ROTATION / SCREEN SUPPORT FLOW BLOCKAGE (SIX LEGS) PLATE Figure 76. Nozzle and plenum chamber design to Figure 75. 15% AV-8B model nozzle air-feed design insure a jet with uniform velocity profile (Ref. 8).

(Ref. 47).

o z 120 1200 NOZZLE EXIT AREA, % o 80 [:3 100 0 120 60 600 4OO .5- 20 200 .4- 0 0 .3- .2- E .8 0 .6 .4

t'!

.2 45 50 0 5 10 15 20 25 30 35 40 Ibf i

2;o

o 5'0 lOO 15o

T,N Figure 77. Effect of exit area ratio on ejector performance (Ref. 98).

PRIMARY AIR SUPPLY INDUCED FLOW PRESSURE RAKE (a) Turbojet (Ref. 97).

SECTION A-A

i.+. c+t_

SIDE VIEW PRESSURE "I =_--+ ,. +,:, RAKE (a) Nacelle details.

_ ==_ _"-_._GAS-GENERATOR FAN FAN FAN AND PLENUM EJECTOR GAS-GENERATOR PLENUM PRESSURES

A*-I

B-B C-C D-D (b) Turbofan (Ref. 99).

Figure 78. Power simulation by ejectors.

(b) Model installed in the Langley 4- by 7-m tunnel.

Figure 80. Use of tip-driven fan in a lifting- nacelle configuration (Ref. 101).

Figure 79. Tip-driven fan assembly.

FAN NOZZLE EXIT AREA = 141.41 (21.92) PRIMARY NOZZLE EXIT AREA = 33.74 (5.23) ,._----15.38 (6.054)-----_--9.5 (3.741)-_ _'--10.98 (4.324)'-_' H.P. AIR SUPPLY_ | TOTAL

_- _.___]_ I PROBES \l I I S'rAT,CS

1 r

.... PRIMARY _ ,_ TECH DEVELOPMENT INC PLENUM (1.316) _- VIEW LOOKING FORWARD 13.97 (5.5) FAN ASSY _ DIMENSIONS: cm (in.)

VIEW ON 'A'--'A' Figure 81. Use of 5.5-in. tip-driven fan in Boeing lift-cruise fan assembly (Ref. 102).

..: ... .: ...,:...:...

........_........ .'.:..:. :.;; _. _ .

• =================================================== ....

::::::::::::::::::::::::: <_>,_:'.:.::::.:.:.:.._ • ._: : ::::::::.:_'.; ....:::.:.:.: Figure 82. 11% Grumman V/STOL model in groun_ Figure 83. CMAPS before installation intQ a model.

effect tests.

iiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiii_i_i_iiiiiiiiiiiiiiiiiiiiiiiiiiiii_iiiiiiiiiiiiiiiii_i_i!iiiiiiiii!_iiii!iiiiiiii_!_i_i_iiiii#i_iiiiii!_ii!i!i_i!i!iiiiiiiii_iiiiiiiii!iiiiiiiiiiiii_iiiii_i!iiiii!i_ii!_!iii_ii!_i!!i!!iii!iiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiii!iiii _:_:_:_:_:_:_:_:_:_:_:_:_:_:.:.:.:.:9:.:.:.:.:.:.:.:.:.:_:_:_:_:_:_:_:_:_:_:_:_:_:.:.:_:_:_:_:_:_:_:_:.:_:_:_:9:.:.:.:.:.:.:.:.:.:.:.:.:.:_:*:_:_:_:_:_:_:_:_:_:_:_:_:9:_:.:_:_:_:_:i:_:_:_.:_:_:_:.:.:.:_:.:_:_:_:_:_:_:_:_:_:_:_:_:9:_:_:_:_:_:_:_:_:_:_:_:9:_:_:_:9:::::_:_:_:::9:.:.:+:.:+:.:+:+:.:_:_:_:_:_:_:_:+:+_:_:+:9:+:+:+:+:.:_:*:_:_:_:_ Figure 84, Schematic of CMAPS.

\

AFT METRICBREAK INLET DUCT CMAPS; JET EFFECTS PLENUM OR NONMETRIC (METRIC) FLOW THROUGH DUCT (NONMETRIC) _] METRIC (ALBEN SHOWN) [] METRIC (SECTIONED) DUCT METRIC BREAK TASK 2.5 IN BALANCE _::::::::: i:+: ....

STRUT METRIC BREAK _ SUPPORT SYSTEM Figure 85. Support system for the CMAPS-powered model shown in Fig. 45 (Ref. 105).

PRESSURE AIR LINES f VACUUM TANK ENGINE INLET AIR PLENUMS VIEW PORT TWIN ENGINED VSTOL WIND TUNNEL MODEL ENGINE EXHAUST TO 1"3 STING EXTRACTORS VACUUM PUMPS 12' DIA METRIC FRAME AXIAL LOAD CELL FLOW CONTROL ISOLATION FRAME VALVING

/

TANK HEAD HINGEDFORACCESS Figure 86. CMAPS calibration tank (Ref. 106).

THRUST W PT/Po A e N (Ib) kg/sec (Ib/sec) m 2 (in. 21 J85-5 11,560 (2600) 19.28 (42.5) 2.24 .0742 (115) 2.1 .0188 (29.2) T58 2,670 (600) 5.67 (12.5) J97 3.0 .0839 (130) 20,020 (4500) 31.75 (70) JT15-D 1.35 .1419 (220} 8,450 (1900) 34.02 (75)

....... / L

--I ---i I

L..... ,I

L--:--=_:.=- ...................... ... I

• -,, ....... "_:2 ...........

J85-5 T58 ......... J97 _.l JT15-D Figure 87. Relative engine sizes of small gas turbines (Ref. 4).

Figure 89. Large-scale upper-surface model using the JT15D engines (Ref. 15).

N Ib _t_] TF34 36 x 103 8 x 103 J52 (1)

+/

6 /

24 P I _ / J97 /

5Pl / / /"

_ 20_ / / , _-JFE731 (t) ._/°

.f"

16 1-

- / ./

121- 2 t'_ : ./J AUG. WING 81-

IT / /J_/S_COMPRESSOR.._"'_-

II • Y T / .,- i 41- 1 I_ /,,/ /.. VIPER

R

0l.,- 1.0 1.4 1.8 2.2 2.6 3.0 3.4 3.8 PRESSURE RATIO (a) 3/4 front view.

REFERENCE LINES Cd - C v • A N • PN "V2/ WHERE: A N = m2 in. 2 0.485 .... 0.102 220 ..... 0.051 110 ..... 0.014 30 Cd = .96 C v = .95 (t) ESTIMATED (b) 3/4 rear view.

Figure 90. Thrust as a function of pressure for Figure 88. A JT15D installation for an upper- several candidate large-scale model engines (Ref. 4).

surface blowing model (Ref. i04).

Figure 92. Nacelle components for the 11%-scale Figure 91. 11% G698 on-board instrumentation shown G698 model.

ready for installation of the fuselage shell.

Figure 93. Drive-air force-isolation coil for the Figure 94. Model parts for the high-speed, two- 11% G698 model, engine fighter model of Fig. 43.

(a) Metal fabrication of model shown in Fig, 8. (b) Soft construction use of polyurethane foam.

Figure 95. Boiler-plate model construction.

INLETAND u NEWSTRAKES U NACELLE Figure 96. Boiler-plate plan for the full-scale AV-8B model (Ref. 8).

AIR O USB ('74) (2 ACTUATORS) CIRCULATION I-I L/C MODEL 1 (2 ACTUATORS) TIME // RANGE IN AUGMENTOR WING (5 ACTUATORS) • _ / SETTLING TIME _// FOR 17,800 N NO REMOTE __% (4000 Ib) THRUST CONTROLS __ CHANGE

r-_L

TURBOJET _=._._

-_._j_:_EFFECT,VE

uJ i- POWER ..Q.... --" .=, 40 ON _ --"

EFOR.'_"__ _q

z / u_ q 268 N/m 2 J--_- = -_'-'- ..... T ....

o

- <>

uJ REMOTELY EVALUATED WITHOUT _ 134 N/m 2 ._POWER CONTROLLED i- MODEL NSTALLED _(2.8 Ib/ft 2) z OFF I 1 I I I I I I I I I I I I o 20 40 60 80 100 120- 2 4 6 8 10 12 14 16 r_ 0 knots NO. OF CONFIGURATION CHANGES I i I I I I 5 10 20 30 40 50 Ib/ft 2 HARRIER V = 120 knots FUEL FLOW PER HOUR 4-JT15's 3719 kg (8200 Ib) V = 120 knots O USB ('74) T = 71,170 N (16,000 Ib) FUEL FLOW PER HOUR [] L/C MODEL 1 30 F / 1451 kg (3200 Ib) L/C MODEL 2 - A HARRIER | / T = 35,580 N (8000 Ib) .E E I_ QSRA

_3

< 2 i- < SCALE READING + c_ OR _ CHANGE + 20 sec I I I I 200 400 600 800 T-,-. , 5-, NO. OF MODEL INSTRUMENTATION READINGS 0 10 20 30 40 50 60 70 80 *ACTUAL TIME PF_R DATA POINT DURING TYPICAL MINUTES FROM TUNNEL START POLAR OR YAWLER Figure 97.

Factors influencin_ testing time for powered models in the Ames 40 by 80 (Ref. 4).

AIRDUCT > OPPOSING MASS FLOW BELLOWS DISTRIBUTER (NONMETRIC) NONMETRIC SEALS (METRIC) __.___

/ .......... X_ _..._._ _////_//_ I V/////7//7//_.\\\\\\\\\\\\\\\\\\\\

--_-->-- IT

I IIIIIIIIIIIIIIIIIIII ...... A,R FLOW RFLOW

JIIIIIIIIIIIIIIIIIIIF//_ _- " )-IWI _ i £

MODEL PLENUM _ "-_1 _\/ STING

(METRIC) _ ! BALANCE FLEXTURES (NONMETRIC) MODEL (METRIC) I

" I

METRIC Figure 98. Schematic of flow-through internal balance (Ref. I09).

DISPLAYS AND RECORDING "--_ DISPLAYS

I RECORDER tt t

REALTIME EXECUTIVE

S..A,.OAO. I CONDITIONING

PROCESSOR JUNCTION HIGH SPEED DATA ACQUISITION SYSTEM BOARDS SAMPLE _ I SIGNAL I_ TEMP I l, ADC CONDITIONINGJ I

t

HOLD 11_ OBS ANALYSIS MODEL __ DYNAMIC F (PCM) PATCH SYSTEM PANEL

I

DATA

I

RECORDING TSP

J

DATA PCM RACK GATHERING

_ N__AL'_' SV±TEM %RLCM PROCESSOR

IL:I I I DISPLAYS

_ CONTROL , : D SYSTEM I TOLEDO

I I c°Ns°LE -I

TUNNEL souND ] NOTAT,ON: AUDIO JB I _I AUDIO JB

CONSOLE OBS ON.OARD SYSTEM

_CM PULSE CODE MODULE

== MOVIE CAMERAS ADC ANALOG TO DIGITAL l TV AND TSP TWISTED SHIELDED PAIR CONTROL Figure 99. 40 by 80 data-acquisition system.

727 FLIGHT DATA ------ STATIC DATA CORRECTED TO FLIGHT _--- 40 BY 80/STATIC DATA J= ¢- 10| .J Z ¢1=

I I I I I I

lOO 90 110 130 150 ANGLE RELATIVE TO INLET AXIS, deg Figure 101. Comparison of 40- by 80-ft wind tunnel Figure 100. JT8D mounted in wind tunnel for the and flight-stand data with the 20-1obe ejector jet noise investigation (Ref. 114).

suppressor on a JT8D-17 turbofan engine (Ref. 114).

J

m-

z mm UJ 90 > <{ t-- O

1'-,_15.4 m---"'l

mr t-- e = 12.3 7O MEASUREMENT OASPL O WIND TUNNEL 121.8 A FLIGHT 120.1 I I 6O SPEED = 40 m/sec (131 ft/sec) 2.87 m BPF = 83 Hz _{ (113 in.) _1 WIND TUNNEL OASPL = 114.8 I STOL BYPASS DUCT I JTI5D-1 = .

I_---HYBRID_ -_)-I_Ew ITH ACOUSTIC_,J

:=

/INLET /TURBOFANI TREATMENT / J

lOO

ACOUSTIC_I'r I tFAN ROTOR I TREATMENT _BYPASS STATOR PRIMARY / z DUCT 9O I- Q ,_ 8O I-.

MICROPHONE UNDER NOSE 7O O FLYOVER DATA A WIND TUNNEL DATA HIGH PRESSURE I I I 6O COMPRESSOR 50 100 1000 10,000 SUPPORT STRUT FREQUENCY, Hz Figure 102. Comparison of corrected flyover and Figure 103. Hybrid inlet nacelle installation (Ref. 122).

40- by 80-ft wind tunnel noise data at equal distance for two aircraft (Ref. 120).

Figure 104. Oil-smear evaluation of flow on multiairfoil two-dimensional model (Ref. 49).

(a) (b) Model at _ = 24 ° , Cj = 2, q = 479 N/m 2 (10 Ib/ft2).

Figure 105. Tuft grid study for evaluating upwash near inlet of EBF transport model (Ref. 130).

0° SOLDER _3 STAINLESS TUBING O. n.

CROSS-SECTION 7 HOLE PROBE (REF. 131)

///////////////////////////

o

_o

INSTRUMENTATION_ AMPLIFIERS I I LPS-II I VIEW A VIEW B

t

STATIC PRESSURE 0.635-cm diam. PORTS 0.71-mm diam.

I PDPll/, I

5 HOLE PROBE (REF. 47) (b) Seven-hole probe (Ref. 131).

(a) Five-hole probe (Ref. 47).

Figure 106. Five- and seven-hole directional probes.

M M _ IF LA 4 WATT ARGON-ION LASER PR POLARIZATION ROTATOR DP DISPERSION PRISM SP SEPARATION PRISM WP WEDGE PRISM BS BEAM SPLITTER BC BRAGG CELL M MIRROR RM ROTATING MIRROR --0.5145#m GREEN DM DICHROIC MIRROR ........... O.4880#m BLUE PT PHOTOMULTIPLIER TUBE .................... 0.4765#m VIOLET IF PINHOLE AND INTERFERENCE FILTER (a) Three-dimensiona] laser velocimeter. Uses three different colors from a single argon-ion laser to form three independent dual-scatter backscatter channels.

-- _ Z DRIVE MOTOR /'--- 7' x 10' WIND TUNNEL _ ,,,,_LT DR,VE ,,, VEW, __ OPTICAL WINDOW ! BALL 3D TILTING LENS_ J_- JSCREW SYSTEM

_r

TILTING./r'("_ MIRROR ¢:::::1 v _ _ _LUE BEAMS VIOLET BEAMS 2"74_'_--_ ° ASSEMBLY I _ _ rE- F _ ____ _ MEASURING LDV PACKAGE _ _ _ / VOLUME S!S_TE _ _EN BEAMS _Z L GROUND PLANE ......._1_!/> Y __ 2D ENS ",,_, JET TRANSLATION PLATFORM I _ i "l PLENUM I ' I L

X-DRIVEMOTOR L_ __1

K- -1

_-J _BEARING -_ I LINEAR I_ FAIRING

I I I

I I I I I (b) Three-dimensional LDAtranslation platform and tunnel installation for the jet-in-a-cross-flow test.

Figure 107. Three-dimensional laser velocimeter (Ref. 132).

7O AMES 40 x 80 ft WIND TUNNEL

l

ROTATION__

_x,_o_/"__'--_ AI ...... I

B V.LOC,.ET.. /

Figure 108. Three-dimensional laser velocimeter Figure 109. Use of two-dimensional laser veloci- operating in the Ames 7 by 10 wind tunnel; jet-in- meter in large wind tunnel (40 by 80 shown) cross-flow test.

(Ref. 135).

Figure 110. Large laser velocimeter in operation.

1 FOLDING MIRROR 2 OUTPUT LENS 3 INTERMEDIATE LENS 13 FOCUSING LENS 4 TRANSLATING LENS MODULE 14 PINHOLE AND INTERFERENCE FILTER 5 FOCUSING STEPPER MOTOR 15 PHOTOMULTIPLIER TUBE 6 MIRROR 16 LASER POWER SUPPLY 7 RHOMBOID ROTATOR ASSEMBLY 10 ABSOLUTE ENCODER 17 LINEAR BEARING 8 BEAMSPLITTERANDFREQUENCYSHIFTER 11 GEAR REDUCER 18 TRANSLATION STAGE 9 LASER 12 ANGLE ROTATION STEPPER MOTOR 19 LATERAL TRANSLATION STEPPER MOTOR Figure 111. Laser-velocimeter system shown mounted on lateral traversing rails; insert shows alternative method of receiving for extended-range configuration (both beam pairs are shown in the insert (Ref. 135).

1. Report No. 2. Government Accession No. 3, Recipient's Catalog No.

NASA TM-85936

4. Title and Subtitle 5. Report Date

May 1984

V/STOL WIND-TUNNEL TESTING

6. Performing Organization Code 7. Author(s) 8. Performing Organization Report No.

A-9694

David G. Koenig

10. Work Unit No.

9, _rforming Organi_tion Name and Addre=

T.3514

Ames Research Center

11. Contract or Grant No.

Moffett Field, CA 94035

13. Type of Report and Period Covered 12. Sponsoring Agency Name and Address

Technical Memorandum

National Aeronautics and Space Administration

14. Sponsoring Agency Code

Washington, DC 20546

505-43-01

15. Supplementary Notes Point of Contact:

David G. Koenig, Ames Research Center, MS 247-I, Moffett

Field, CA 94035 (415) 965-5047 or FTS 448-5047

16. Abstract

Factors influencing effective program planning for V/STOL wind-tunnel

testing are discussed. The planning sequence itself, which includes a short

checklist of considerations that could enhance the value of the tests, is

also described. Each of the considerations, choice of wind tunnel, type of

model installation, model development and test operations, is discussed, and

examples of appropriate past and current V/STOL test programs are provided.

A short survey of the moderate to large subsonic wind tunnels is followed by

a review of several model installations, from two-dimensional to large-scale

models of complete aircraft configurations. Model sizing, power simulation,

and planning are treated, including three areas in test operations: data-

acquisition systems, acoustic measurements in wind tunnels, and flow surveying

18. Distribution Statement 17. Key Wor_ (Suggest_ by Author(s))

Wind-tunnel testing

Unlimited

Powered-lift testing

Propulsion simulation

V/STOL aircraft

Subject Category - Ol

19, Security Classif. (of this report) 20. Security Classif. (of this page) 21. No. of Pages 22. Dice °

74 A04

Unclassified

Uncl ass i fied

"For sale by the NationatTechnical Information Service, Springfield, Virginia 22161