Click for next page ( 16


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 15
STATUS OF HYPERSONIC TECHNOLOGIES 13 2.0 TECHNOLOGIES RELEVANT TO HYPERSONIC VEHICLES AND THEIR STATUS 2.1 Aerodynamic - Propulsive Integration As usually conceived, the hypersonic air-breathing propulsion system uses a low speed system for operation from standstill to about Mach number 2.5, a ramjet for operation to about Mach number 6.5, and a scramjet for Mach numbers above 6.5. If these three systems are to be combined{, this must be done in a way that does not dlegrade their individual performances when they are active, and with acceptable weight and complexity. This is a major chal- lenge for the designer of hypersonic aircraft, and must be recognized as such. We will not deal with it comprehensively in this discussion, however. Our focus here is primarily on hypersonic propul- sion. Integration of the airframe and propulsion system is a central feature of all conceptual designs for hypersonic flight vehicles mainly because the engine capture area must be a large fraction of the vehicle frontal area. Among the contributing factors are: 1) a low thrust per unit of engine air- flow at hypersonic speeds, which results from the small fractional change in energy of the engine air- flow that can be achieved through combustion; 2) the need to fly as high as possible to minimize the heat load on the structure, which results in a pro- portionately low engine mass flow for any given capture area; and 3 ~ the desire to make efficient use of the compression by the bow shock of the vehicle, which leads to the need to capture, in the engine, most of the flow through this shock. The need to maintain a weak bow shock to minimize losses and for a large capture area require slender configur- ations in which the entire forebody, or at least its lower surface, comprises the engine inlet. (See Figure 2-1.) The same factors dictate that the aft end of the fuselage serve as the expansion sur- face for the propulsive streamtube. While the resulting configurations are conceptually appealing, especially to the propulsion-oriented, they pose prob- lems that, though not entirely new, are certainly more serious than for more conventional designs, in which the pro- pulsive streamtube and fuselage and wing airflows are farther apart. Thus, in most if not all conceptual designs for hypersonic vehicles, the propulsion system is assumed to ingest the boun- dary layer flow that develops on the forebody. Most successful propulsion installations in the past have avoided this. If the propulsion system ingests the boundary layer: l ~ the ramjet, whether operating in the subsonic or supersonic combustion mode, must pass two parallel streams of very different velocities and temperatures, or 2) the boundary layer flow must mix with the free-stream. The former will lead to constraints on the supersonic combustion process, because the pressure must be equalized between the supersonic and subsonic streams, imposing serious performance penalties. The latter may result in losses, or in heating of the supersonic stream, which is counter to the principal rationale for the supersonic combustion ramjet, namely to lower the combustion temperature. These issues are discussed further in the propulsion sections. However, it is

OCR for page 15
14 proper here to ask whether the engine should ingest the boundary layer from the forebody. One can imagine con- figurations in which the boundary layer flow bypasses the combustor, the inlet ingesting flow primarily from outside the boundary layer, but we have seen little evidence of serious consideration of this possibility. The high degree of integration also poses serious control problems if, as seems probable, the inlet compression and nozzle expansion occur only on the lower surface of the vehicle. Both the inlet and nozzle then contribute strongly to the vehicle's lift, particularly at high flight Mach numbers. A balance be- tween the lift forces on the inlet and nozzle will determine the pitching moment produced by the propulsive streamtube. While the lift and thrust can conceivably be oriented through the center of mass and the pitching moment can be nutted at the design point, it is not yet clear how these balances can be maintained at off-design conditions, without large forces from control sur- faces. Furthermore, even if a design can be evolved to meet these require- meIlts in normal operation, what pitching moment will follow from a sudden loss of heat addition in the combustor, and the resulting modification of the nozzle expansion? Although an inlet upstart probably will be unacceptable in a scramjet operating at very high Mach numbers, how will control deal with such an upstart? A further control difficulty may arise from the ingestion of the ramp boundary layer by the engine. To see this, suppose the airplane pitches upward so as to increase the angle of attack. The ramp boundary layer and shock layer now accumulate more losses, higher static pressure, hotter and lower stagnation pressure air, and these are ingested into the engine. With no change in fuel flow, tile engine thrust and the pressure on the nozzle face HYPERSONIC TECHNOLOGY FOR MILITARY APPLICATION change and ~ strong pitching moment is developed. Although quasi-steady operation of the engine is involved, the coupling with the airframe dynamics is strong and must be dealt with by the control system. This is not only an issue of sensors and control response, it is a question of engine and airframe design. For example, should steps be taken to design or position the engine to be less sensitive to such disturb- ances? In the extreme, should the engine avoid ingesting the boundary layer entirely? The presently proposed configurations for the NASP are subject to such airframe interactions that may pose significant control problems. Most of the current airplane config- urations use a modular propulsion system with several engines side by side under the airframe. The individual modules probably will operate under nearly identical conditions and usually not interact or interfere with each other. However, when changing propulsion mode from low speed engine to ramjet and from ramjet to scramjet, each of the modules will sometimes experience a start-up transition that might induce airflow disturbances that propagate upstream of the inlet. It is unlikely that this phase of operation will occur simultaneously for each of the modules and, indeed, it may not occur symmet- rically with respect to the desired thrust axis. Furthermore, if for some reason the starting transients are severe, the inlet disturbances or inlet malfunction can propagate from one module to the other, perhaps leading to a catastrophic malfunction of the propulsion system. Such difficulties can be closely coupled with yawing disturbances of the airplane which may, in turn, be induced by unsymmetric mode changes of the engine modules. The large nozzle expansion required for efficient hypersonic operation leads to ~ serio us base drag problems at transonic speeds, where the nozzle

OCR for page 15

OCR for page 15

OCR for page 15
STATUS OF HYPERSONIC TECHNOLOGIES pressure ratio is far too low to fill the entire base area. This problem has appeared in non-afterburning operation of aircraft where the nozzle area required for afterburning operation has dictated a large base area. Ejector nozzles, which fill this base area with a secondary or tertiary air stream, have provided partial solutions, but this avenue will be much harder to follow for hypersonic aircraft. Base burning is a possible alternative, but one whose heating and fuel consumption implic- ations have not been fully explored. Another class of problems arises from the need to integrate the low speed propulsion system with the high speed ramjets, with acceptable weight and complexity, and without serious interference with the function of the scramjet at high Mach numbers. There should be no extraneous projections into the airflow that would cause strong shocks or excessive heating. All sur- faces of the engine flow path must be actively cooled, and this further argues for a minimum of complexity. While we certainly do not argue that innovative solutions to these design problems are improbable, there is no base of success- ful designs on which to draw to solve them. The importance of aerodynamic- propuisive integration for hypersonic vehicles has been recognized for many years, and has been highlighted over the last two by active participants in the NASP program and by advisory groups. But today there is more enthusiasm than integration. This problem is not being adequately addressed and will remain so as long as responsibility for the pro- pulsion system and vehicle are divided. We urge that an organizational structure be created" where responsibility for the conceptual design of both engine arid vehicle resid e in one organization. This must be done soon, so that these issues are faced in the conceptual design 15 phase, not after a configuration has been selected. 2.2 Propulsion Systems In addition to the several technol- ogical areas that are either unique to the scramjet or are emphasized to an unusual degree, there are important issues of ( 1 ) transition (2) the limit- ations to development and testing above Mach number 8, and (3) the extraordin- arily sensitive interaction between the engines and the airframe. These issues will be discussed separately after we have examined the basic technologies pertinent to the high-speed engine. 2.2.1 Basic Scramjet Engine If the requirements underlying the principles fundamental to the scramjet engine are not satisfactorily met, the engine may perform~no better, or per- haps worse, than a high performance rocket. The main issue is to maintain the static temperature of the air in the combustor at a reasonable value while the aircraft is flying in the Mach num- ber range 10-24. At a very high air temperature the eventual reaction of hydrogen and oxygen to water vapor is very slow or incomplete or both, and the specific impulse of the engine falls from a value in excess of 1000 seconds to well below 500 seconds. This concept breeds two conflicting technological problems. First, due to the high air velocity in the combustor, the combustor would have to be very long to achieve a reasonable residence time. Second, extreme heat transfer rates and wall shear losses require the combustor to be as short as possible. How to balance these issues and whether or not there is an acceptable balance underlie the factors discussed below. The processes of injecting the hydrogen fuel and mixing it with air in

OCR for page 15
16 the scramjet appear to be the most dif- ficult obstacles to the realization of a successful engine; and they are processes in which our present fundamental and technological base is weakest. Hydrogen fuel must be injected into the engine with very low losses, at local Mach numbers as high as 8, where losses have the greatest effect because of the small fractional heat addition due to combus- tion. It is generally agreed that shear layer mixing rates drop under some conditions of supersonic relative motion between streams. The technological basis for this is not extensive and design experience is lacking. Possible alternatives, such as mixing augmenters, wall injection, and shock enhanced . . . mixing are In an even more primitive technological state. It must be made clear here that satisfactory mixing for a chemical reaction process contrasts sharply with one in which, for example, momentum is being exchanged. The mixing must be complete on the molecular level to allow combustion. Not only is this a more time-consuming process but the exper- imental difficulties of assessing the completeness of molecular mixing are considerable, and therefore the tech- nological basis will be slow to develop. Considerable effort is now being expended in appropriate investigations and the results will be of unusual value not only to the present development but also to future efforts of scramjet development. It is not now clear just how extensive the data will have to be to impact the NASP Program. During the most important periods of scramJet operation the combustor Mach number is in the range of 2.5 to S. This flow field is quite complex due to the heat release, which is controlled by the molecular mixing process, and by the ramp boundary layer and bow shock layer that may be ingested by the engine. The heat release has a pro HYPERSONIC TECHNOLOGY FOR MILITARY APPLICATION pounced effect on the structure of the flow field which, in turn, strongly influences the mixing processes. This coupling introduces complexities that we find very difficult to cope with either experimentally or computationally. Today it is impossible to describe with certainty the best geometry of the com- bustion chamber, for any Mach number or altitude. A serious concern among workers experienced in the field is the stability of the hypersonic flow in the combustion chamber during combustion. Would a small disturbance imposed on a design flow field decay, diverge, or lead to a pulsating combustion process? A central obstacle to understanding the result of such a time-dependent disturbance to the combustion chamber flow is our current incapability to either experimentally measure or to compute this chemically- reacting flow. The ingestion of the ramp boundary layer and shock layer may lead to large variations in the air density and stag- nation pressure over the cross-section of the inlet. The air density from top to bottom of the engine inlet may some- times vary by a factor of four. Exper- ience has shown that such conditions facilitate communication of disturbances through the boundary layer ahead of the engine, which may result in unfavorable inlet conditions and even inlet instabil- ity. Such a disturbance may couple with the combustion process in the chamber with possibly unfortunate results. Our experience with this problem is restric- ted to a much lower Mach number re- gime than that appropriate to the NASP and requires serious experimental atten- tion and perhaps design compromises. Almost exclusively, our ability tO calculate chemically-reacting unsteady flow fields is restricted to one dimen- sion. Such steady one-d imensiona1 calculations are being used in engine performance calculations, yet quasi one

OCR for page 15
STATUS OF HYPERSONIC TECHNOLOGIES dimensional analysis can cope neither with the mixing-controlled combustion issue nor with the stability problem. But difficulties arise even at the more elementary level of performance calcu- lation. When the engine ingests the ramp boundary layer and bow shock lay- er, the gas entering the combustor has a very non-uniform temperature distribu- tion over its cross-section. As a con- sequence, the chemistry, which has a strong and non-linear temperature dependence, may vary even more vio- lently over the cross-section. In trying to adapt one-dimensional analysis to this problem one is faced with the issue of choosing appropriate average values for each cross-section, which introduces large and unacceptable errors in the results. Also, the output of a one- dimensional analysis can provide only a uniform input to the nozzle calculation that follows. Even an approximate cal- culation of the nozzle expansion process makes it clear that the pressure distrib- ution over the nozzle surface and, con- sequently, the calculated performance of the engine, is very sensitive to the details of the gas state distribution at the start of expansion. It is most important to make at least an approx- imate accounting for the two- and three-dimensional nature of the combus- tor and nozzle gas dynamics. As a result of the relatively high temperature and short residence time in the combustion chamber, the hydrogen- oxygen reaction will not reach an equilibrium water vapor concentration before reaching the nozzle. So it is important that the reaction between OH and H be completed to the maximum degree during expansion in the nozzle. Because it is most unlikely that crucial experiments can be done, a high degree of reliance must be placed upon compu- tations. These must, at the least, be for two-dimensional reactive flow arid pref- erably three-dimensional. Moreover, we must account for the non-uniform state of the gas at the nozzle entrance. At present this magnitude of numerical calculation is impossible and the situation is unlikely to change very soon. The nozzle may be one aspect of the NASP that undergoes significant development during flight test. As we have noted above, the ingestion of the ramp boundary layer and shock layer exacerbates the problems of the intake, combustor, and nozzle. Although it is difficult to study without better computational capabilities, we must nevertheless consider whether or not the boundary layer should be swallowed. Even at the cost of some possible reduction of the airplane performance, the additional design certainty that would accrue from capturing clean air might result in an overall benefit. 2.2.2 Cooling Problems Although active cooling with hydro- gen will be employed over several sections of the airplane, the larger part of the cooling load is connected with the engine. Even the most optimistic estimates show that the hydrogen flow rate required for cooling exceeds the stoichiometric fuel flow requirement above Mach number 15. Less optimistic analyses suggest that above Mach num- ber 20 the hydrogen cooling requirement may exceed! four times that for combus- tion. This means that the molar flow rate of hydrogen in the cooling passages of the airframe is more than twice the total flow rate of air into the engines. And because it is the molar flow rate of gas that determines the pumping power requirement, details of the cooling passages and the active management of the coolant flow to different regions of the airframe and engine are as impor- tant as the external gas dynamics. In fact, because the airframe is largely sized by the hydrogen tankage, the cooling requirement exerts an unusually high leverage on the airplane size and

OCR for page 15
18 weight. The actual cooling requirement is now one of the least certain elements in the entire vehicle, so the airframe probably will be over-sized to accom- modate this uncertainty. The design of cooling passages in the engine and assurance of their effectiveness are made much more dif- ficult, and may perhaps be compromised, by the extensive geometric changes required of the engine over its Mach number range. It is in the regions of the hinged joints, and the scramjet may require several of them, that meticulous design and careful coolant management must be exercised. The emergence of coolant effective- ness, active coolant management, and a high degree of integration of the coolant system with the structural design as a significant and novel aspect of the NASP is just beginning to be taken seriously. Furthermore, this issue will command attention for any hypersonic airplane, and the technology base that is acquired here will be of permanent value. 2.2.3 Propulsion System Transition The acceleration of the NASP from take-off to orbital velocities requires not only widely different operating conditions for each engine but the transition between the two or three different engines, each of which is designed to operate in a specific Mach number range. In the "conventional" arrangement, one engine will operate from zero to low supersonic Mach num- bers, a ramjet with subsonic combustion up to the lower limit of the scramjet range and, finally, the scramjet to near orbital velocities. It may be anticipated that these changes of propulsion mode .. . . . . wit he sensitive processes, requiring careful control to assure that each newly fired engine starts properly and that the transition induces no extreme and unusual loads. Though we have HYPERSONIC TECHNOLOGY FOR MILITARY APPLICATION some experience with this sort of problem, two important factors make the present situation unique. First, the transition to the scramjet mode lies in a Mach number range far outside our experience and where an error may be costly. Second, in contrast with our experience, experimental verification and development will not be possible. The chance for serious trouble here is so likely and the potential damage so great that the issue is significant. The development problems of the individual propulsion systems are so severe in themselves that considerations of this problem might be postponed too far into the program. 2.2.4 Auxiliary Rocket Propulsion Because the ground-based test facilities for the NASP will be unable to accommodate the Mach number and enthalpy range appropriate to the high speed engine, much of the engine testing and development must be planned as part of the flight program. Engine develop- ment and testing will occur as the flight envelope of the airplane is expanded into the higher Mach number range. During this process, the airplane will require a separate propulsion system to allow a range of conditions under which to test the engine. This probably would take the form of several hydrogen- oxygen rockets distributed appropriately over the airframe but physically separate and with an independent control system. These rockets should be capable of taking over the propulsion and trim of the airplane during testing of the high- speed engine. This will allow much greater flexibility in exploring the operating envelope of the engine at nearly constant inlet conditions. This requirement will hold not only for the NASP but for any hypersonic research vehicle using scramjet propulsion. It appears that there is a quite separate requirement for some degree of

OCR for page 15
STATUS OF HYPERSONIC TECHNOLOGIES rocket propulsion as a component of the final NASP propulsion system itself. Rocket propulsion will be needed to insert and stabilize the vehicle in orbit, and possibly to maneuver in orbit and de-orbit. It is very likely that the rockets will prove invaluable in main- taining thrust and trim during change of propulsion mode. Part of the long-term growth pro- cess for a conventional aircraft and propulsion system is the gradual improvement of the thrust level and efficiency of the engine, and a cor- responding improvement of the overall aircraft performance. The NASP will be no exception to this pattern and will, in fact, undergo even more striking changes than the conventional aircraft. In its initial form, the scramjet probably will become ineffective at velocities con- siderably below orbital velocity and will require a component of rocket propulsion during some final portion of its accel- eration. But the long-term growth of the engine will reduce, though probably not eliminate, this requirement, and the propulsion system must accommodate this growth without severe design changes to the entire airplane. Rocket thrust should be an integral part of the final operational NASP propulsion, not just a temporary feature of the flight test program. 2.3 Aerodynamics The aerodynamics of a hypersonic vehicle must be closely integrated with the power plant. Presentations to this committee arbitrarily focused on the aerodynamics of the entire vehicle including the inlet, but excluded the propulsion system element, consisting of the fuel injectors, combustion chambers, and expansion nozzle, despite the fact that many items included in the aero- dynamic discussion are important in the propulsion system itself. The aero- dynamic discussions must address flight 19 conditions from take-off to orbit, conditions of high dynamic pressure (2000 psf seems to be an approximate, practical, upper limit), low Reynolds number conditions at high altitude, and conditions in the non-continuum region approaching free molecular flow. The aerodynamic considerations require the treatment of real gases with full viscous effects applied to configurations that consist of blunt noses and leading edges on slim bodies with complex configur- ations that generate three-dimensional gradients and shock waves. The requirement is to predict, with reason- able accuracy, the following parameters, assuming similar inputs will be provided by the propulsion system: Local pressures, heat transfer, and skin friction, on the surface of the entire vehicle, including the flow through the inlet. a) To provide the forces (lift, drag, moment, etc.) over the full flight envelope b) To provide the detailed thermal loads for cooling and structural design, and c) To provide dynamic inputs to the aerodynamic and thermal . control systems. The three-dimensional flowfield detailed information on the conditions and the state of the with local gas, for configuration and control design, for inlet positioning conditions, and optimum inlet design for flow to the combustion chamber. The statuses of the various aero- dynamic technology elements have been evaluated on the basis of the experi- mental data base from wind tunnels and flight, and from the examination of computations that have been validated or calibrated under the appropriate condi- tions. Current computational capability enables us to calculate complex flows, but is limited in dealing with the details of viscous arid real gas effects. Current technology provides the capability to

OCR for page 15
20 compute real gas effects if the signif- icant reaction rates are known. In general, aerodynamic characteristics for three-dimensional configurations up to about Mach number 10 are reasonably known or can be predicted, if transition location and the extent of transition can be provided. Beyond that, where real gas effects become important and viscous effects more complex, very little three- dimensional aerodynamic information is available beyond the blunt body con- figuration for realistic design procedures. The following comments address the inclusion of detailed viscous effects, real gas effects, and real gas complex flow conditions. 2.3.1 Viscous Effects The primary viscous effects are experienced in the boundary layer where the hypersonic conditions require a considerable extension of low speed experience and computations to high Mach numbers and "cold wall" conditions. The main problem is to define the condition of the boundary layer at any point on the vehicle during any cond- ition of flight. This definition is required to calculate heat transfer and skin friction (critical elements in the design of the structure and cooling system), to predict and avoid unaccept- able separations, and to compute per- formance. 2.3.1.1 Transition Point Determination Determination of the transition point is a critical element in the design and performance of a hypersonic vehicle. Unfortunately, there is no well-founded theory for this determination. Conven- tional low Mach number use of the para- meter Rex/M at values as low as 150 to about 300 has not been validated under the conditions of hypersonic flight or for complex flows. New attempts to use the parameter en are being studied by HYPERSONIC TECHNOLOGY FOR MILITARY APPLICATION NASAiLangley. The problem is that wind tunnel data are biased by wind tunnel disturbances and free flight results are primarily for simple bodies, axisymmetric, with zero gradients. Considerable work is being done in this field, including extensions of laminar, linear stability theory, and attempts to evaluate cold wall effects in the lower hypersonic region. Still missing are high Mach number data with 3-D gradients and shock waves. Proposals are being considered for extended wind tunnel tests under the appropriate conditions and for flight tests of bodies that might include more complex geometries. The problem is made even more difficult by the lack of knowledge of the disturbance field in the stratosphere, which could be a factor in triggering transition. 2.3.1.2 Extent of the Transition Region Although it is critical to determine the transition point, the condition of the boundary layer downstream of that point is also important. There is some indi- cation that the transition region between fully laminar and fully turbulent flow gets longer as Mach number increases. It is possible, therefore, that a con- siderable region of transitional flow will occur over a hypersonic vehicle or in a hypersonic inlet. The characteristics in this region are unknown. It is believed to be partially laminar and partially turbulent, in a temporal sense, but detailed information in this region must await solution of the transition point problem of sub-section 2.3.1.1. 2.3.1.3 Turbulent Boundary Layer Somewhere downstream of the tran- sition point, the flow will be fully turbulent. At higher Reynolds number conditions, low altitude flights ~ considerable part of the hypersonic vehicle may be turbulent. Depending on the trajectory chosen, this turbulent

OCR for page 15
STATUS OF HYPERSONIC TECHNOLOGIES flow could occur under Mach number conditions that have not been exten- sively studied. The effects of high external Mach number and cold wall, with gradients have not been well- val~dated for the higher Mach number regimes. 2.3.1.4 Test Conditions It is possible to do considerable work on items 2.3.1.1 through 2.3.1.3 in ground facilities below Mach number 10, although the results to date are only preliminary and effects of wind tunnel conditions on such phenomena as the transition point have not been fully explored. The effects of cold walls in 3-D complex flows, with the gradients and shock waves that are required for optimized configurations in this lower Mach number range, along with inlet studies, can be studied, but have not. These same problems at high Mach numbers are critical issues because facilities are limited, and all of the parameters for boundary layer validation must be applied, including wind tunnel disturbance effects (now including entropy and constituent effects), wall cooling and catalysis, and wall rough- ness. 2.3.2 Real Gas Effects Blunt body flows have been calcula- ted and validated to some considerable detail. There is some question of the results for very high altitudes where the constituents of the atmosphere and the local conditions are not well known. Accurate measurement of the conditions ahead of the vehicle in flight would provide an important input into the analysis of the results, and considerable effort is being expended in this area' although solutions to the problem are not yet in hand. 21 Slim, complex, three-dimensional bodies with blunt roses and leading edges provide a special problem in real gas effects. Although the blunt nose solutions are well known, the streamlines downstream of the blunt nose carry different histories; mixing and recombin- ation rates depend on local conditions. A similar problem is experienced in the expansion of the combustion chamber flows through the nozzle. Although the chemical kinetics for the blunt nose are well known, not all of the data needed for recombination rates for the down- stream flows are in hard. These data should be obtainable, however. The effects of real gas kinetics on boundary layer characteristics are only beginning to be explored. The effects are probably not important below Mach number 10. Above 10, surface catalysis and chemical reactions in the boundary layer flow, with cold walls, could affect boundary layer characteristics and the transition point and transition region. Studies are only in their elementary phase and some reasonable solutions for the boundary layer characteristics will be required before the addition of real gas effects can be attempted in detail. 2.3.3 Real Gas Effects in Complex Flows Inlets, bodies' fins, and wings, and their interactions, cause a combination of viscous effects, shock waves, and strong gradients with real gas effects. These must be understood to give detailed flowfields, such as for the combustor, where the local conditions and the constituents (state of the gas) must be known at each point. Local, configuration-specific hot spots must be identified. Inlets have been tested at low Mach number. At high Mach num- bers, the inlet tests nest be closely associated with ~ ~ en -defined initial flow determined by the details of the forebody. The lack of adequate testing facilities and the ~or~plexity of the flow

OCR for page 15
48 sure. Unfortunately, the Navier-Stokes equations have long been known to be very inaccurate for computing the flow structure within shock waves, so these equations cannot be used to get reliable CFD results for the particular conditions existing on cowl lips in high altitude flight. Since the shock on cowl lip heating may be very high, it must be investigated more realistically than it has been, by either direct simulation Monte Cario methods, or by appropriate new experiments, or by continuum equations more appropriate than Navier- Stokes. 2.7.1.3 Combustion Flows Computational fluid dynamics codes for scramjet combustors in hypersonic flight are in a relatively primitive state because of two circumstances. First, models for turbulent mixing in reacting compressible flows have been far less successful to date than have models for boundary layer flows. Second, experi- mental data on hydrogen-air mixing in scramjet combustors for flight faster than Mach number ~ have not been available to provide any code calibration in this important Mach number range. The degree of incomplete fuel-air mix- ing, and the nonuniform distribution of both species concentration and of ther- modynamic state are issues of potentially vital importance. It is unfortunate that the turbulent-mixing and chemically reacting type of flow in combustors, which is yet to be investigated exper- imentally for flight above Mach number 8, is also the same type of flow for which present CFD computations are the most uncertain. 2.7.1.4 Nozzle Flows In the expansion of combustion products through a nozzle, or a partly wall-bounded nozzle, reaction-rate chemistry is essential to the flow HYPERSONIC TECHNOLOGY FOR MILITARY APPLICATION computation. Currently 2-D reaction- rate codes are used to compute nozzle exhaust expanding into ambient air, but 3-D reaction-rate codes for an expansion partly over the surface of a vehicle, and partly into a vehicle-dependent external aerodynamic flow, do not yet exist, although they are under development. Vital to a nozzle flow computation is knowledge of the distribution of species, thermodynamic state, and dynamic quantities exhausting from the combustor. One must also know what chemical reactions are important, and what their rates are, since non-equil- ibrium chemistry is essential in nozzle expansion. In the nozzle flow compu- tation, a new element arises because the exhaust-fuselage boundary layer might relaminarize. Like transition, relaminar- ization is poorly understood for the conditions of hypersonic flight. 2.7.2 Validation of Hypersonic CFD Codes An experimental validation of non- equilibrium hypersonic CFD codes is more difficult than for conventional aircraft codes because of the absence of ground-based test facilities that can stimulate together the total variety of physics represented in hypersonic codes. Different experimental facilities, however, can test different components of the overall hypersonic physics simulated in the codes. Hypersonic wind tunnels, for example, can test a code's ability to simulate perfect-gas flows over complex 3-D geometries, although not always at the desired flight Mach and Reynolds numbers. Shock-tube type facilities, on the other hand, can test a code's ability to simulate high temper- ature thermochemical aspects of a flow, although usually for simplified geomet- ries and not always at the desired Reynolds number. Even though ground facilities cannot test together all interacting components of the physics

OCR for page 15
STATUS OF HYPERSONIC TECHNOLOGIES represented in a hypersonic CFD code, it is nevertheless essential to test by comparison with experiment as many of a code's physics components as is feasible. Even when this is done, some aspects of a hypersonic code (e.g. transition location, and nonequilibrium radiative heating) cannot be adequately tested by comparison with available ground test experiments. Flight tests may be the only way to thoroughly test the ability of a code to accurately compute such flow field parameters. 2.7.3 Future Role of CFD Overall, hypersonic CFD today appears acceptable for external aero- dynamics and inlet flows provided that the location of transition is known, and for nozzle flows, provided the initial entrance conditions are known. CFD, however, is weak for combustor flows, and is unable to reliably predict the location of transition. We must recog- nize that these current limitations are not inherent to CFD, but are mainly a consequence of the present state of supercomputer development which forces the use of a Reynolds-averaged form of the Navier-Stokes equations. While this . tlme-averaglng process pro( uces equa- tions that are solvable on current computers in a practical amount of time, it requires that the turbulent stresses and heat flux be modeled, thereby introducing the primary inaccuracies in present day codes. If the full time- dependent Navier-Stokes equations were employed instead, the turbulent eddies would be directly computed rather than modeled, and the degree of CFD realism would be expected to greatly increase. Such calculations for a complex three- dimensional vehicle are outside the reach of today's supercomputers. However, they may be feasible for computing the onset and extent of boundary layer transition on a fuselage using current or next generation supercomputers. This advanced type of computation would require the development of methods for computing through transition in a hypersonic boundary layer on a vehicle flying through a prescribed disturbance field. Knowledge of the disturbance field in the stratosphere would, of course, also be needed. These obstacles to the direct numerical simulation of transition and turbulence (DNST) prob- ably will be overcome in the long range future. This will open up an entirely new level of CFD capability having much greater realism than present capabilities provide. The potential future importance of DNST to the Air Force is great, since it would largely overcome most of the present limitations of CFD. Such a capability would provide a large increase in the effectiveness of CFD applications to the design of aircraft and turbine engines as well as hypersonic vehicles. In view of such a major future potential, the technology of DNST computation is clearly important to the long range interests of the Air Force. Since this potential would serve industry and other agencies as well as the Air Force, it may by itself justify the development of future supercomputers with power suf- ficient to realize this next-generation level of advanced CFD technology. 2.X Experimental Capabilities Ground test requirements for hyper- sonic flight vehicles, even for cases of simulation rather than duplication of flight conditions, impose extreme demands on the equipment in terms of pressure and temperature (see Figure 2- 8~. Further, such facilities are expen- sive, ranging from $1-2 million to in excess of $500 million and take several years to construct or bring on-line. During the 1 960's extensive hypersonic test facilities were constructed in the U.S., and overseas, so that by 1971, 52 major operational aerodynamic test units existed, as shown in Figure 2-9. How

OCR for page 15
so ever, during the 1 970's and l 980's the number of operational facilities was reduced dramatically so that by 1986 only 23 were still in useful condition (see Figure 2-10~. Engine test facilities are similarly restricted (see Figure 2-11~. Indeed, while recent interest has led to some moth-balled facilities being refurbished some are still being consid- ered for destruction. Further, many of these facilities are 20 to 40 years old and do not produce appropriate flow conditions to address the problems presented by hypersonic air-breathing vehicles and their develop- ment. Indeed, recent studies of the upper atmosphere indicate that the com- mon concept of a quiescent and rela- tively ur~iform regime (a goal of some facility developments such as the NASA Quiet Tunnel) may not represent reality. Also, research is needed on facilities and their effect in particular on such phen- omena as boundary layer transition and chemical kinetics. Even where ground testing has been used extensively, such as the space shuttle with 35,000 occupancy hours and other vehicles (see Table 2-C), flight performance is not always well pre- d icted. For the hypersonic regime, extrapolation to portions of the flight envelope will further increase the probability of error. CFD codes will help in some cases; however, code validation in many areas remains to be done and requires the same ground test facilities discussed above and in the following sections. Thus, for the foreseeable future, it appears that it will not be possible to verify advances in many of the hyper- sonic technologies through ground testing. Consequently, flight test programs are needed for components of hypersonic air-breathing vehicles that cannot be adequately tested in ground facilities, for validation of concepts, and proof of CFD codes. Such testing HYPERSONIC TECHNOLOGY FOR MILITARY APPLICATION represents an extension of the ground test concept and is complementary to, not a substitute for complete systems ground testing. Experimental aircraft such as the projected NASP research vehicle will be essential in expanding the data base for large portions of the flight envelope. In the following sections general facility requirements, and then specific requirements for aerodynamics, propulsion, and materials and structures, will be discussed separately. 2.~.1 Test Requirements Test requirements for a hypersonic vehicle capable of flying over a speed range up to orbital (Mach number 25) are extremely severe. Data for aero- dynamic, propulsion, and structural materials are required up to the very high pressures and temperatures at which air becomes a hot plasma com- posed of molecular and atomic particles, ions, and free electrons. In stagnation regions of a hypersonic vehicle at high altitudes, atmospheric oxygen begins to dissociate above about Mach number 7. Both oxygen and nitrogen are almost fully dissociated at Mach number 15, and ionization becomes important at Mach numbers approaching 20. Data are required for pressure distributions, surface friction, temper- ature distributions, heat transfer rates from cooled surfaces, chemical reaction rates at high temperatures, and in locally disturbed flow, fuel/air mixing rates, material properties at high temperatures, and structural response. The key test parameters for aero- dynamic testing are Mach number, Rey- nolds number, Knudsen's number, and facility total enthalpy as well as transients in these variables. For the combustion processes associated with the type of air-breathing cycles being considered, this list is substantially

OCR for page 15
~ ~- ~ - o o o o o o o IN A) v) a, E It r _ ~ O ~LL - ~V :3 O e_ ' r\' . en O _ O O ~ 0 ~ HO 0 ~ C O I]~] ]0 SONYSOOH] `]OnlIl]V O q_ ~ 0 So _ . a_ _ Or cy ~ ~3 O _ - ~: o _ _ ~ _ o s - LL - - , C

OCR for page 15
\\ - - ~; // ~ / I/- D. ~\ Cal ~ o (D Cal Cal hilt .,~. = l - i` ~,,~/ ~ , 11 1 1 i,, I 1 11, , , 1,1 , , ~ o o o o a' o 5 - CI: Z ~ Cat OC C_ In 3 a_ _ Cal e_ U) ~ ' O ~ ~ i: 3 3 Go CD Cal 1

OCR for page 15
- ' a- ~ at ~ ~ / /l ~ r:v rim / v K1 HIT Li7 Zip 1 V ~ V \/ 1 1 , ~ ~ 1 ~-- 1 ' 1 1 1 Or 1 ~= 3 .' L I I I I I r l of ~ O O O O In \_ a _ O ~_ a) c em Z Cat \ /\/ a' - O ~ Cat ~ 1 . ~ C. - Cot CD Let, - a: o Sat ._ - ._ Cut E Lot - Cot to U) o ~ 3 - a:

OCR for page 15
-\ /-~-. ~ - ~ 5) - o \ \ I: - v ~ v ~ ~ ~ - ~ - - - ~ Ed k 1 ,. i ! 1 ' . 1 1 ~: - 4,, ~ ~ ~ In Ct (I;) A ~ in, ._ Cd ~ I:: ~N~ 1 1 ~-_ lo o o ~o Cal Cal o o Us o En Cal ~ ~_ ,4 , - o - _ D Q I_ Cal 1 Cal Cal A: o i_ a_ Lo In Lo CO a,

OCR for page 15
as o ~ >) ' Ad of ~ Ad ~ 0 ~in a, ~tD toto ~LOof Al en it: ~Q ~ E ~ in O _ ~ a' 0Cal ~0 >, E ~e' O ~ Oo a) Q ~Cal : O ~O _ A) ~ O) ~1 tat ~._ Cat _ ~o oO ~ O. O ~c ~ ~3 ~ ._ O :~: _ _ I ~1LO ~ C ~m. O ~ ~ au ~coc ~co C~ Q s~ a) ~0 C~ _ ~'_ ~o~ oC~C~0 ~' ~)e ee t ~_1 ~C~ t_~_ =), 1 >, O 3r ~ C~.O C'~ S~OC~l w)~- ~QI II =:X C~L~

OCR for page 15

OCR for page 15
STATUS OF HYPERSONIC TECHNOLOGIES longer and allows for less flexibility in terms of simulation through dimension- less variables. Thus, it is necessary to consider Prandtl number, Stanton number, Eckert number, Lewis number, and species chemical reaction rates non- dimensionalized by residence time. For some cases scaling or non-dimension- alization is ineffective and full-scale hardware testing is needed, such as with structural panels and joint sections. (See Appendix C: Glossary and Appendix D: Dimensionless Groups in Fluid Mech- anics for explanations of these dimen- sionless variables.) From an aerodynamic and propulsion standpoint it is necessary to locate the region of boundary layer transition on a slender vehicle forebody to determine flow conditions in the inlet of air- breathing engines and the subsequent combustion process, particularly above Mach number 6 for which scramjet pro- pulsion is envisaged. Thus, full Reynolds number and Mach number should be reproduced in a ground test facility, if full simulation were to be achieved. In contrast to a rocket-powered vehicle such as the space shuttle, which can have a steep ascent trajectory, a hypersonic vehicle using air-breathing propulsion for boost to orbital velocity would require a trajectory over an ex- tended Mach number range in the lower atmosphere to meet the air mass flow requirements of its engine. This implies a high dynamic pressure of about 1500 psf corresponding to which the Reynolds number on a full-scale vehicle would exceed 100 million up to Mach 10 or higher. Although a steeper rate of ascent probably would have to be fol- lowed above Mach number 12 because of temperature limitations on materials, Reynolds numbers of the order of 10 to 20 million can be expected up to above Mach number 20. For free stream Mach numbers above about 10, test requirements become even 51 more severe for slender vehicles because Mach number, free flight Reynolds num- ber, and full enthalpy must be repro- duced in a single test facility as real gas effects become important. In con- trast, a blunt vehicle such a the Apollo capsule or even the space shuttle (which reenters and stays at a high angle of attack - 40 deg - down to about Mach number 10) requires full enthalpy sim- ulation, but not the full Mach number and Reynolds number. Since viscous and high temperature effects are important for virtually all hypersonic testing, facilities using air as a medium are required because other gas media (freon, helium, pure nitrogen) do not provide the right quantitative simulation and there are no satisfactory means of converting all the required data to air. The above requirements hold for both aerodynamic/aerothermodynamic and propulsion testing. For the latter, it is not sufficient to test at free stream conditions corresponding to the lower Mach number at the immediate engine inlet (direct connect testing) because of flow distortion in the actual flight inlet due to ingestion of the thick forebody boundary layer, the impingement of shocks generated upstream and shock wave-boundary layer interactions. Thus full Mach number and forebody geometry are required to test the propulsion system. Another important use of hypersonic ground test facilities will be validation of CFD codes as the dependence on numerical techniques as a design tool is expected to be extensive in the hyper- sonic regime because of physical limit- ations of wind tunnels at the higher Mach numbers. This will require wind tunnels with capabilities not only at the desired test conditions, for example Mach number and Reynolds number for aerodynamic

OCR for page 15
52 tests, but with well-documented flow quality due to their effect on important phenomena such as boundary layer tran- sition. Most of the existing tunnels were built primarily for lift and drag type measurements or leading edge heat transfer and do not have good enough flow qualities for code validation. Thus, in general, facilities using air as the test medium and operating Rey- nolds numbers of 100 million would be required for full testing of a hypersonic cruise vehicle intended for operation up to Mach numbers 10 or 12, while Reynolds number up to 10 to 20 million and Mach numbers up to 25 would be required for an air-breathing orbital vehicle. Furthermore, full temperature simulation and high Re would be required for a hypersonic lifting vehicle operating above about Mach number 10. 2.~.2 Aerodynamic Test Capabilities Facilities for aerodynamic and pro- pulsion testing in the subsonic, tran- sonic, and supersonic regimes (below Mach number 5) are adequate to meet most future requirements if some facil- ities in need of repair are rehabilitated. Above Mach number 5 there are some 30 major facilities 1 in the United States and eight in Western Europe, the United Kingdom, and Japan. All but one half dozen were built in the 1 950s and 1 960s. All of the newer facilities were built in the early to mid- 1 970s. None are known to have been built after 1 976.2 Of these 30 facilities, seven (vir- tually all in U.S. industry) are on standby status, i.e., they are not pres- ently operational. Only seven hyper- sonic facilities using air are capable of yielding Reynolds numbers based on chord lengths in excess of 20 x 1 o6. (See Table 2-D.) HYPERSONIC TECHNOLOGY FOR MILITARY APPLICATION Of these high Re facilities only the Calspan shock tunnels are able to pro- duce free flight total temperatures above Mach 10, and only the 96-in. shock tun- nel approaches total temperatures for a Mach number well in the teens. Test durations for these facilities are a few milliseconds. From a propulsion standpoint, scramjet research has been undertaken in the 4 ft. diameter Scramjet Test Facility at the NASA Langley Research Center at Mach number 6 and temper- atures up to 4000 R on small models under one sq. ft. in cross-section. Scramjet tests up to Mach number 7 were run in the HRE hypersonic test facility at Plum Brook, Ohio, over a decade ago, but this facility has not ~ seen in use since. Aside from limited parameter simu- lation, a drawback in most existing facilities is that the flow quality is not good enough for boundary layer transi- tion simulation. Boundary layer transi- tion on wind tunnel models has generally been found to occur at much lower Reynolds numbers than in free flight. This is due mainly to disturbances in the flow emanating from wind tunnel settling chambers and acoustic radiation from nozzle wall boundary layers. A super- sonic facility designed to minimize such disturbances has been under study at the NASA Langley Research Center - the "quiet supersonic tunnel". Present plans are for a Mach number 3.5 capability with possible addition of a Mach number 6 nozzle later. Thus, capabilities for aerodynamic and propulsion testing to meet require- ments in the hypersonic regime are extremely limited below Mach number 10 and virtually non-existent above Mach number 10. Because wind tunnels (even shock tunnels despite their very short running times) are temperature-limited for