CORRUGATED AND COMPOSITE NOZZLE-INLETS FOR THRUST AND NOISE BENEFITS . -._ ,- .. M. Glinsky Hanapton University, Hampeom, VA 23668 1.M. Blankson NASA G k mR esearch Center, Clevekmd, OH 44135 V.G. Gromov and i VJ. Sakharov c)2001 American Institute of Aeronautics 8 Astronautics or Published with Permission of Author@) andlor Author@)' Sponsoring Organization. CORRUGATED AND COMPOSITE NOmEI"S FOR ._. TFIRUST AND NOISE BENEFITS M. Gmhslry* Hampton Universitj, Hamptou, VA 23668 IM.B lanlrron** NASA GlmR esearch Center Cleveland, OH 44135 V.G. Gmmov*** and v.L sllkhlvov**** Institute of Mechanics at Moscow State Univemity Moscow, 117192, Russia ABSTRACT . - ThZ following research results are based on development of an approach previousfy proposed and investigated in [14] for . - optimum nozzle design to obtain maximum thrust The design was W0ted.a Telescope nozzle. A TeIescope node contains one or several mkmal designs, which an inserted at certain IoCations hto a divergent Conical or planar main nozzle near its exit. Such a design provides additional thrust augmentation over 20% by comparison with the optimum single nozzle of equivalent lateral area, What is more, experimental acoustic tests have discovered an essential noise reduction due to application of Telescope nozzles. In this paper, some additid theoretical results arc presented for Telescope nozzles and a similar appro& is applied for aemperfommce improvement of a supersonic inlet. Numerical simulations wcrt conducted for supersonic flow into the divergent portion of a 2D or axiqmmetric node with s e v d plane or conical designs as well as into a 2D or axisymmefric supmnic inlet with a forebody. The Kryko-Godunov marching numerical scheme for inviscid supersonic flows was used. Several cases were tested using the NASA CFUd and IM/MSU Russian codes based on the full Navier-Stokes equations. Numerical simulations were conducted for non reacting flows (both codes) as well as hr real high temperatme gas flows with non-equilibrium chemical reactions (the latter code). In general, these simulations have confinned essential benefits of Telescope design applications in propulsion system. Some preliminary numerical simulations of several typical &et designs were conducted with the goal of inlet design optimization for maneuvering night conditions. I. INTRODUCTION research of Ahuja, Krothapalli et al. [5,6] has shown Several well-known experimeatal results show that inserting disturbing elements into supersonic jet essential acoustic benefits in the application of some flow: slots, finger, tabs etc., can reduce jet noise (and untraditional nozzle designs. For example, nozzles screech tones) in spite of the presence of numerous with rectapgular or elliptic cross section in the strong and weak shock waves. This contradicts the supersonic part produce less jet noise than round traditional view of the considered phenomenon. A nozzle designed for a fixed Mach number at the reasonable explanation for these facts would be the nozzle exit ( i.e. with uniform flow at the exit and appearance of more effective mixbg and deatruction pressure coinciding with the flight static pressure of the regular cell-shock structure in the weakly outside the exhausting jet). Thus, the theoretical underexpanded jet. Inside such a jet, weak barrel- perfectly shock fiec jets are "noisiex" than at least shaped shock waves are always present and these partially underexpanded (or overexpand@ jets with shock waves are the main sources of the oscillatory possible internal shocks. Moreover, the experimental processes in the jet. In the regular almost parallel co- annular mixing layers, unstable longitudinal waves * -Research Professor, Senior Member AIAA are excited, and noise is produced in a fixed direction ** Senior Scientist, Associate FeUow AIAA tkom the jet axis -145'. Of come, the presence of *** -Leadingscientist shock waves in the jet exhaust, especially for a ****-Leading scientist supersonic nozzle, can lead to some dangerous side effects and performance penalties. Copyright 43 2001 by heA merican Institute of Aeronaubics and AstrOIliIuticsInc. Aurightsreserved , c)2001 American Institute of Aeronautics &Astronautics or Published with Permission of Author(s) andlor Author@)'S ponsoring Organization. 2 Developing previous ideas for jet noise reduction, external nozzle so that the integral of the pressure on two novel concepts were proposed in the papers [I- the low side of the inserted surface is greater than on 41. The first concept is denoted as the Bluebell the upper side produces increased thrust. There is an nozzle, based on the flower-like shape of its external optimal angle of the plate that provides the maximum jet plume. Bluebell nozzles utilize both chevrons and thrust at each point of a divergent flow. The most corrugation in their nozzle geometry. An example of efficient internal design is produced from a pattern such a design is shown in Figure 1. The second that looks like a telescope with extending tubes. The concept is denoted as the Telescope nozzle for it optimal number of internal designs is defined through consists of several internal nozzle surfaces that are dependence on the Mach number at the nozzle exit, arranged in a telescope fashion. Each concept is M, Telescoping designs must be located so that the cap&,!e of XEelmkig a Lhu-it peflL--u-IlnllC--U-l bC greater coqressib!e waves hmed by ktemct;,m of a flow than the standard baseline conic or 2D plane with this design would be passed on to the upper side convergent and convergent-divergent (CD) nozzles. of the next lower telescoping part. The best result Several modifications of such designs are shown in will be produced by such a set if the external design Figures 2-4. The improved performance of Bluebell inclination increases downstream. Computations nozzles occurs due to the increase in nozzle internal show that a significant thrust benefit from the surface area while maintaining nozzle-projected &ea Telescope nozzle occurs with an external telescoping equivalent to the baseline reference nozzle. Small design, using either wedge, conical or optimal scale and large scale acoustic tests of different contour shapes, and also in the case of a plug modifications of Bluebell nozzles were conducted at application. the NASA Langley Research Center and the Central AeroHydrodynamics Institute (TsAGI) in Moscow, 2.2 Optimum plate location. In the usual case, a Russia. These tests have shown essential acoustic prate (or airfoil) inserted into an inviscid supersonic benefits of Bluebell design applications in supersonic flow produces a resulting force normal to the plate regimes as well as in subsonic regimes. For example, whose magnitude and direction depend on the the experimental tests of several Bluebell nozzle pressure difference on both sides of the plate. The designs ([ 31) have shown noise reduction relative to a non-dimensional aerodynamic characteristics of the CD round nozzle with design exhaust Mach number plate, the thrust, T, or drag, CD, produced by this m=1.5. The best design provides an acoustic benefit flow about the plate can be calculated with these four near 4dB with about 1 % thrust augmentation Below, parameters: specific heat ratio, K, flow Mach number, we consider only the second (Telescope) concept an angle a between the flow and thrust direction, with the goal of design optimization for the and the angle y, between the flow and the plate. The maximum nozzle thrust for the intended application angle, y, is measured from the upstream flow of this concept to propulsion systems, especially, for direction. If P O a nd less than a limiting angle ybm, a supersonic engine inlet and in stationary detonation (OSspm), the thrust (drag) is determined by the engines. Detailed information about the first simple analytical formulae using relationships for (Bluebell) concept is in the papers [ 1,2,4] and in the oblique shock waves and the Prandtl-Meyer patent [7]. Some information about the second rarefaction wave discussed in the previous sections. (Telescope) concept is in the paper [2] and in the In this interval of the angle y, for all other parameters patent [ 81. there is an optimal value of the angle yqt, which gives the thrust maximum value. Aerodynamic XI. PLATE ELEMENT IN SUPERSONIC FLOW characteristics of the unit plane element in supersonic flow were calculated using the created code for a 2.1 The thrust on a plate element with an oblique shock wave and Prandtl-Meyer rarefaction flow. wide range of the parameters: K, M, and u. These calculations have shown that there are some optimal A divergent flow can act on a plate or airfoil inserted parameter values which provide the maximum thrust. into a flow so that a resulting force is directed against Note, that for large attack angles to the thrust the flow. This effect is used for thrust by supersonic nozzles. Conversely, a uniform flow produces only direction, a, maximum thrust is obtained at the drag for bodies and airfoils. Inserting a cclnical or limited angles, nh. wedge-shaped nozzle inside the divergent part of an 42001 American Institute of Aeronautics & Astronautics or Published with Permission of Author@) and/or Author($)' Sponsoring Organization. 3 Similar results were observed for another case that was essentially improved for the considered problem corresponds to a pure Pmdtl-Meyer flow at the solution. A cylindrical (x, r, (p) coordinate system for turning point of the 2D nozzle wall. Again, for small axisymmetric and 3D cases or Cartesian (x, y, z) angles, a, there are maximum thrust values inside the coordinate system for 2D case with components of a interval OIsvma,n d for greater 0: maximum thrust velocity vector q on these axes of (u,v,w) is occurs at the limiting value yt,i,,. considered, and define q to be the modulus of the velocity vector q, p is the pressure, and p is the 111. NUMERICAL METHODS density. All variables are non-dimensional. Linear sizes are related to a throat height Y, for 2D problems 3.1 Theoretical approaches. The general purpose of or radius r. for the axisymmetric subsonic nozzle rhe theoreticai approach is to defie the optimum portion. For the iniet problem, characteristic hear conditions what provide minimum nozzle thrust loss sizes were nonnalized by inlet entrance values. or maximal thrust augmentation by comparison with Velocities were related to the sound velocity c. in the the baseline convergent-divergent design or conical nozzle critical section (throat), density by the critical nozzles. To achieve an optimal nozzle design, the density p., and pressure by p. e?. The gas is solution would require multiple computations of a 3- assumed perfect with constant specific heat D supersonic flow region. For practical applications, coefficients c, and c,,, so that the specific heat ratio Reynolds numbers Re are very high -106-108. Thus, 1c= cp /c, is constant. The Euler equation is written in the boundary layer at the wall is turbulent and makes the form of the well-known integral conservation up a small portion (-1-3%) of the cross section area laws. and to an even lesser extent affects the longitudinal nozzle size. The grids for these problem solutions were constructed during flow calculation downstream For fast preliminary numerical analysis of the inside some sectored domain between two meridional investigated problems in such a situation, it is planes of symmetry for 3D problems, and along the inefficient to use numerical solutions based on the normal direction to the nozzlejet axis of symmetry full unsteady Navier-Stokes equations. Our approach or plane of symmetry for axisymmetric and 2D flows was based on the "viscous-inviscid interaction'' [3]. respectively. We omit any explanation of grid We used the Euler approximation for definition of the generation and all finite relationships for this scheme. "external" inviscid flow outside a thin boundary Description of them can be found in the paper [3] and layer with fiction along the nozzle wall. The Euler in the original Russian book [9]. calculations were repeated for each new nozzle shape after accounting for the boundary layer thickness. 3.3 Numerical Methods for Solving the Navier- The new computed "external" inviscid flow again Stokes (Russian MMSU approach). The two- was used for definition of a new boundary layer dimensional Navier-Stokes equations are solved in thickness. In each iteration, of course, the boundary curvilinear structured mesh through a finite volume layer is computed at the original nozzle surface. approach. Under this approach, the finite difference Usually, the results were closed after several equations system consists of numerical analogies of iterations, between 3 and 4. However, the latest the conservation equations for quadrilateral cells numerical simulations have shown that correction of covering the computation domain and difference the total thrust value for the supersonic nozzle approximation of the boundary conditions; This portion by viscous effects is not essential for the method yields an approximate solution as a set of considered designs. We neglect such corrections in primitive variables in the centre of each cell and in the present numerical simulation results in this paper. the centre of each cell side lying on the wall. The inviscid numerical fluxes through cell sides are 3.2 Krayko-Godunov numerical marching calculated from the result of the exact Riemann scheme. The Kryko-Godunov explicit 1s t order problem solution for the frozen state of the space-marching numerical scheme (K-G-code) [9] considered non-equilibrium thermal and chemical was employed for numerical simulation of supersonic processes. The interfacial values are defined by the flow in the divergent nozzle part and jet exhaust and limited one-dimensional extrapolation primitive in the supersonic flow portion of inlets. This scheme . c)2001 American Institute of Aeronautics 8 Astronautics or Published with Permission of Author(s) and/or Author(s)' Sponsoring Organization. 4 variables from the cell-centres to the cell sides. The increases along the nozzle centerline from zero at the numerical viscous fluxes through cell sides are throat, x-., to the maximum value at the exit, evaluated using the central and one-sided difference k&* x, where x=(x-&)/(x, x.), i.e. this coincides formulas of the second order accuracy. Difference with definition of a corrugation amplitude coefficient equations are solved by the two-layer fully coupled 6. The cavity (convexity) width, A m 91 , also implicit iterative scheme based on implicit linearly increases (decreases) downstream fiom zero approximation of the time-dependent Navier-Stokes (maximum) at the throat to the maximum (zero) at equations. The Gauss-Seidel line space-marching the nozzle exit, i.e. 2cpl=T-A~x. For such a method with the LU-decomposition of the block- configuration, two expanded flows near the nozzle tridiagonal matrices along lines is used on each time wall flow into two neighboring cavities to meet each step. other at some angle a, mutually penetrate and more effectively mix. A flow impulse on the lateral area of The Navier-Stokes solvers are developed for gas- the convexities increases the resulting nozzle thrust. phase models of various complexity from perfect gas to thermally and chemically non-equilibrium models A Chisel nozzle is very convenient to use with a including excitation of internal energy modes, Telescope nozzle as shown in Figure 2 because chemical reactions and ionization due its coupling. similar convexities allow the internal design to be For all classes of models the databases on maintained. In this figure, the main external design is thermodynamic and transport properties of individual based on the cone of the angle -1 00 and the internal chemical species are created. The each realization of design has a conical surface. In Figure 3, the external a model is made out as a set of files containing nozzle is constructed by giving the fixed contour information about the chemical gas composition and z'z(x) in the zx-plane and a cross section contour is kinetic processes included in the given model. By the described by the super-elliptical equation: special programs-generators this information is transformed to data sets needed for compilation and (y/a(x>)"'"+(Z/b(x))"'"'=1, (*) run of the flowfield calculation codes. Gas-phase models can be used with various gas-wall interaction where n(x)=2+H(x-x,~*~(x-x.)/(~-x,), models. In the boundary conditions on a body surface c(x)=a(x)/b(x)=lCH(x- x,)*cdx- x*)/(&- x.). the slip effects, finite rate of exchanges by the internal energy modes, heterogeneous catalysis and The Heaviside function H(x-x, is defined ablation can be taken into account. This approach H(x-x.)=O if x,$ x S x., and H(x-x,)=l if x. Ix % was described more detail in the papers [ 10,113. *, The subscripted indices 0, and e correspond respectively to the nozzle inlet, throat and exit. The IV. TELESCOPE NOZZLE NUMERICAL subsonic portion of the nozzle (from the inlet to the SIMULATION RESULTS throat) has axisymmetric shape (a=b=l, n<). In the supersonic portion (from the throat to the nozzle 4.1 Telescope nozzle geometry. Examples of two exit), the power n in (*) changes from the minimal possible Telescope nozzle embodiments are shown in throat value of 2 to the maximal exit value &and the Figure 2 and 3. In Figure 2, the external design is a eccentricity c=a/b changes from the minimal throat Chisel nozzle. This nozzle can be constructed on the value of 1 to the maximal exit value ce. The nozzle base of any plain nozzle. For the simplest design in contour z=f(x), in the plane of symmetry, y=O, is a Figure 2, dependence of the radius on the azimuthal cubic parabola in the subsonic part and then becomes angle in the cross section is described by a periodic rectilinear with the angle =loo. For the Telescope function r=r(cp) with a period T=2d&: in the first nozzle in Figure 3, the power n(x) increases from 2 to period r=r+=const for OlcpscPl and (p2 S@T, and r=r. 10 downstream from the throat to the exit and two =const for <PI IqP<(p2,w here cpl=O.5(T-A<~)a nd (p2 plane internal designs are located symmetrically =OS(T+A(p). We call a corrugated surface part a supported by holders into the external design. "cavity" or a "convexity relative to the intemal normal to the nozzle wall. The cavity depth (or The Bluebell nozzle concept for jet noise reduction convexity height) defined by the equality Ar=r+-r., can be also used in the Telescope nozzle for either . c)ZOOl American Institute of Aeronautics &Astronautics or Published with Permission of Author(s) and/or Author@)' Sponsoring Organization. 5 the external design or the internal design or both. mount to 100%. This value can be even larger Such an example is shown in Figure 4. There a 6- depending on the angle p. Experimental petal internal conical design is installed into an 8- measurements of small and large scale Telescope petal external design in which three plane holders nozzles are being addressed in future work. These maintain its position. experimental tests are very important because the recent experimental acoustic tests have discovered 4.2 Numerical simulation results based on inviscid essential jet noise reduction produced by nozzles with approximation with the K-G code. The numerical a uniform plate set at the nozzle exit. Such noise simulations were conducted using a modified benefit was predicted in the papers [3,5]. numerical code based on the 1st order explicit ......" uuulerica! marcling scheme of Kryko-Co&iwv 191. haiogous benefits can be obtained €or nozzles with Solutions are obtained using an arbitrary curvilinear a spike (or centerbody). Such an example is coordinate system, and the marching coordinate x is illustrated in Figure 8. Two thin airfoils are located at chosen close to a local streamline. A multi-zone the spike end, x=l, another one is at the jet boundary approach and non-uniform grid application were used inside of the shock layer behind the barrel shock to obtain results of high resolution in complicated wave. Mach contours analysis has shown that there is geometric domains. some optimum location of this airfoil that provides maximum thrust augmentation. These optimum The thrust calculations for the Telescope nozzle parameters are very close to the values estimated on without plug (centerbody) with one to four internal the basis of the analytical approach for supersonic designs have shown that the benefits can be increased flow at the unit plate element discussed earlier and with several internal design applications. Their presented in the paper [4]. In particular, the angle of - location and angles to the thrust direction should be attack of the airfoil to the thrust direction is close to chosen so that each shock wave formed at the lower 10' as in the case shown in Figure 6 for the Prantl- side of the upper design would not intrude upon the Mayer rarefaction wave where this value is close to upper side of the lower design. Similarly, each +loo. Again, thrust augmentation by the Telescope rarefaction wave formed at the upper side of the Spike nozzle application can be obtained up to -50% lower design should not intrude upon the lower side by comparison with the case without additional plane of the upper design. Figures 5 and 6 illustrate this or airfoil designs. This effect will be very efficient approach for three and four internal components for for strongly underexpanded jets (high flight altitude axisymmetric and 2D Telescope nozzles. The thrust conditions). In this case, almost the whole of the jet benefits for conical nozzles can reach up to -30% by exhaust gas is concentrated inside the thin shock comparison with nozzle thrust without internal layer at the jet boundary. The gas density as well as designs. Depending on the optimization conditions, pressure in this layer behind the strong shock wave is this value can be even higher. For hypersonic very high. A plate or solid umbrella-shaped surface nozzles, the augmentation is greater. An example of can be installed in this shock layer, as, for example, numerical simulation results for the 2D wedge- shown in Figure 8, and this umbrella can produce shaped nozzle with four internal thin airfoils is shown additional thrust for the main vehicle. The installation in Figures 6 and 7. Here Mach contours and four problem is outside of our competence. streamlines (black solid lines) are presented. These streamlines correspond to the internal airfoil 4.3. Navier-Stokes Numerical Simulation Results. stagnation points and also represent the zone Several designs were tested using the Russian boundaries. This picture illustrates the essential IM/MSU code based on the full Navier-Stokes benefits of internal design applications inside the equations for two gas-phase models [ lo]. For pure air Prandl-Meyer rarefaction wave and region of nozzle-jet flow, the model of chemically frozen constant parameters at the nozzle wall. In this case, nitrogen-oxygen stoichiometric mixture with the inlet Mach number is K=2, the angle j3=30°, and equilibrium excited of internal degrees of freedom is the thrust augmentation for divergent nozzle portion used. For the premixed hydrogen-air mixture is q=ATK'-75%. Note that the working efficiency of exhausted from a divergent supersonic nozzle to the a Telescope nozzle grows as the inlet Mach number supersonic CO-flow, the simplest non-equilibrium increases. For example, for Ms5, the value of h can model of 7 components Hz, 02,N z, HzO, 0, H, and , c)2001 American Institute of Aeronautics 8 Astronautics or Published with Permission of Author(s) andlor Author(s)’ Sponsoring Organization. 6 OH with 8 chemical reactions is employed. In detached shock wave at the cowl; or d) continuous Figures 9 and 10 the Mach number contours are compressive waves along a curved inlet surface shown inside 2D wedge-shaped supersonic divergent (forebody) with detached shock wave at the cowl. nozzle with inlet Mach number, Mi, = 1 for the Other designs use e) partial external compression and nitrogen-oxygen mixture hydrogen-air mixture internal; f) internal compression; and g) detached respectively. The composition of inlet mixture is shock wave before the inlet. assumed to be chemically equilibrium. These jet flows exhaust to the supersonic air co-flow with Calculations conducted by G.I. Petrov and K. Mach number, Ms2. Analogical pictures are Oswatisch have shown that the total pressure loss in shown in Figures 11 and 12 for inlet Mach number, an ideal inlet may be not more than -3-5% for a Mi, =2. The numbers, iocation and inciination of the shock wave system with equai intensity. In reai inlet internal designs in these tests are the same as shown flow, separation zones can be formed at the sharp in Figure 5 and 6 respectively. change of centerbody inclination or at the point of interaction of the shock wave with the boundary layer Comparison of inviscid approximation results with at the forebody or cowl. This makes the inlet NSE based results for air flows with the same nozzle characteristics worse. Detailed analysis of the configuration shows very close thrust values for supersonic inlet problem is in G.N.Abromovitch’s Reynolds number, Re= 1O 6 -1 O’, where Reynolds book [ 123. Note that most of the optimization theories number is calculated on the basis of critical (sound) and numerical simulation methods for improving parameters. For hydrogen-air flows thrust inlet efficiency do not take these effects into account. augmentation by internal designs a little bit less. Note, that direct application of the Telescope nozzle Separation and inlet drag are important obstacles for with only internal designs inserted did not lead to efficient inlet design. To reduce these effects, effective combustion behind the oblique shock waves application of 3D corrugated surfaces similar to those at the internal designs. Only installation of some that were used for improving nozzle designs may be wedge shaped or spherical obstacle at the plane (for employed. For example, a star shaped forebody or its 2D nozzle) or at the axis of symmetry for the conical smooth modification can reduce forebody drag. Also, nozzle allowed positive and essential thrust it is known from hydrodynamic stability theory (C.C. augmentation. At the present time, this problem is Lin, and others) that 2D velocity distributions in being investigated numerically with the goal of boundary or mixing layers are less stable than optimal parameters definition for a stationary corresponding distributions in 3D cases because there detonation engine nozzle based on the Telescope is one additional degree of freedom for perturbation design amelioration. Semi-empirical separation criteria show the same phenomenon. Several unusual curvilinear IV. INLET NUMERICAL SIMULATIONS surfaces were proposed and tested experimentally many years ago by Russian scientists. However, such 4.1 Supersonic inlet problems. The main purpose of shapes have not been used in the aviation industry a supersonic inlet is to slow down the gas flow from and require further research. Preliminary calculations supersonic speed to low subsonic speed before the are very promising. chamber (compressor). Simultaneously, the total pressure should have minimal loss for effective 4.2 Telescope Inlet Proposal. Most of the combustion in the chamber. The first investigations Untraditio~ln ozzle designs discussed above for and analysis of this problem took place in the 60’s. supersonic nozzles (chapter 111) can be employed for For 2D and axisymmetric inlets, the investigations a supersonic inlet improvement. In particular, the showed that flow total pressure loss through a set of Telescope nozzle and all results of theoretical inclined oblique shock waves with a last normal analysis of this concept are useful. In this case, the shock wave is essentially less than through a unique energy of the turned flow along the forebody wall can detached shock wave before the inlet. Several inlet be used for creation of additional thrust. As in the flow regimes are possible for supersonic inlets. previous analysis, the mutual locations, sizes and These are a) two shock waves at the inlet plus one angles of the internal plates (thin airfoils) are very external at the cowl; b) three plus one; c) three with a important for efficiency of the application. Optimal . c)2001 American Institute of Aeronautics 8, Astronautics or Published with Permission of Author@) and/or Author@)’ Sponsoring Organization. 7 values of geometric parameters were determined simulations results for several other designs similar to from multi-parametric numerical simulations based the duPont inlet confgurations. The goal of these on the modified marching K-G code. The effect of tests is to examine the previous results obtained in four thin airfoils installed at the minimal cross inviscid approximation and for determination of any section (near the corner point) is illustrated in Figure viscous effect’s influence on the main conclusions. 13. Here Mach contours and corresponding These estimations were conducted on mid size streamlines are shown for the 2D Telescope inlet vehicles for low flight altitudes. with a wedged forebody. This design provides a forebody drag reduction of 25%. Obviously, the same Below, we illustrate some designs numerical approach is applicable for other designs, such as simulation with taking into account of viscous p-fis.is~e=ctf;,ci= s -.?&+&crlaess sect;,oE effects. Two fi-meical codes were iwd: the ?USA supersonic tunnels, blunt bodies with several ring- LaRC CFL3D code [13] and Russian IM/MSU code shaped sheets, etc. The star-shaped forebody with 3 [ 10,113. Both codes are based on implicit upwind 2nd ring-shaped pylons is shown in figure 14. The pylon order numerical schemes (EN0v ersions) for solution cross section is a thin airfoil. Its chord inclination is of the full unsteady and steady Navier-Stokes and directed so that it produces thrust augmentation. Euler equations. The examples of such results for These pylons are located in different streamlines of study solutions are shown in Figures 15-18. For the the compressible flow behind the bow shock wave. conditions, free stream Mach number, M, =1,75-3, Therefore, the fuel injected downstream will mix and Reynolds numbers, Re, =I Os - 1O ’, the tests have with air stream uniformly creating premixed flow. been conducted. These results have shown an essential dependence of the flow regime with the 4.3 Optimized Inlet Proposal. In an unpublished design geometry change. For example, a small paper of Anthony A. duPont entitled “Further Studies change geometry can lead to flow transformation of Optimized Inlets for Hypersonic Turbine from the regime with an oblique shock wave fixed on Engines”, the author analyzed several versions of an the cowl top to the regime with a detached shock inlet in the range of flight Mach number, M, ,f rom 2 wave (Fig.10). Figures 11-13 show an influence of to 5. Using simple semi-analytical theory in the boundary layer separation on the wall at the inlet inviscid approximation, he constructed several 2D entrance to the flow for different inlet designs. The inlets including the cowl and wall contours that separation leads to decrease of an effective inlet cross should provide minimal thrust and total pressure section and forms the flow with three bow shocks losses at the compressor (or/and combustor) inlet. configuration. The boundary layer suction from the There is a great interest to realize this approach wall at the inlet entrance allows to eliminate experimentally. One of the problems for the proposed separation and to obtain flow with the oblique shock design is to create an efficient numerical algorithm wave fixed on the cowl top (Fig.18). what can define the inlet geometry corresponding to flight speed, altitude, etc with optimal integral V. CONCLUSION characteristics, such as minimal forebody drag, total pressure and thrust losses. Our previous research was Theoretical analysis and numerical simulation results directed at solution of this problem. We investigated were obtained for nozzles and supersonic inlets with two areas. The first is to check the duPont results the goal of aeroperformance improvement. The using more exact theory, more powerful computers designs investigated are based on development of the and numerical approaches. The second is to approach proposed by the authors for optimum determine inlet characteristics in a wide range of nozzle design for obtaining maximum thrust. Such a flight Mach numbers using a minimal number of design was denoted a Telescope nozzle. A Telescope fmed inlet shapes. The solution of these problems is nozzle contains one or several internal designs at under development. Some results were presented in certain locations in the divergent conical or planar the paper [3] what were obtained using the simplified main design near its exit. Such design provides marching numerical scheme [9]. Comparison of these additional thrust augmentation over 20-30% by results with the prediction theory of A.A. duPont comparison with the optimum single nozzle of the shows essential differences between the two equivalent lateral area. Recent experimental acoustic approaches. In this paper we present some numerical tests have discovered essential noise reduction due to , c)2001 American Institute of Aeronautics 8 Astronautics or Published with Permission of Author@) and/or Author(s)' Sponsoring Organization. 8 Telescope nozzles application as well. Some AIMASMEISAEIASEE Joint Propulsion Conference, additional theoretical results were presented for the 17-19J une, 2000, Huntsville, Al. Telescope nozzle and a similar approach was applied 4. Gilinsky, M., Blankson, I.M., et al., 1999, Aeroperformance and Acoustics of the Nozzle with for aeroperfoxmance improvement of a supersonic Permeable Shell, AIAA Paper #99-1924, 5th AIAAICEAS inlet. At the same time, the classic gas dynamics Aeroacoustics Conference, May 10-12,1999,S eattle, WA. problem of a similar flow at the plate in a supersonic 5. Ahuja K.K., 1993, Mixing Enhancement and Jet Noise flow has been analyzed. In some particular cases, Reduction Through Tabs Plus Ejectors, AIAA Paper 93- new exact analytical solutions were obtained 4347, 15th Amacoustics Conference, Oct. 25-27, for a flow at the wedge with an oblique shock wave. 1993bng Beach, CA. Numerical simulations were conducted for 6. Krothapalli, A. and King C.J., 1993, The Role of supersonuc flow into a divergent portion of 2D, Streamwise Vortices on Sound Generation of a Supersonic axisymmetric and 3D nozzles with several plane, Jet, 15th AIAA Aeroacoustics Conference, Oct. 25-27, conical or corrugated designs as well as into a 2D or 1993bng Beach, CA. 7. Seiner, J.M., and Gilinsky, M., 2000, Undulated Nozzle axisymmetric supersonic inlet with a forebody. The for Enhanced Exit Area Mixing, US Patent #6,082,635. 1st order Kryko-Godunov marching numerical 8. Seiner, J.M., and Gilinsky,M., 1999, Jet Nozzle Having scheme for inviscid supersonic flows was used with Centerbody for Enhanced Exit Area Mixing, US Patent boundary layer correction in thrust calculation #5,924,632. formulae. Several cases were tested using the NASA 9. Godunov, S.K. et al., 1976, Numerical Solution of CFL3d and Russian IM/MSU codes based on the 111 Multidimensional Problems of Gas Dynamics, Moscow: Navier-Stokes equation. The Telescope nozzle Nauka, 1976,400~. concept allows renewed consideration of the 10. Afonina, N.E., Gromov, V.G., and Turchak, L.I., stationary detonation engine concept. Our Numerical Simulation of High-Temperature Viscous preliminary estimates have shown very good Flows, 1999, Proceeding of 7th Annual Conference of the prospects for such engines. In general, numerical CFD So- ciety of Canada, Halifax, May 30-June 1, 1999, pp. 4-3 4-8. simulation results have confirmed essential benefits 11. Gromov, V.G., Sakharov, V.I., and Fateeva, EL, 1999, of Telescope design applications in propulsion Numerical Study of Hypersonic Viscous Chemically systems. Reactive Gas Flow Past Blunt Bodies, Fluid Dynamics, ~34,1999p,p . 755-763. VI. ACKNOWLEDGEMENTS 12. Abromovich, G.N. Applied Gas Dynamics, 1976, Nauka, Moscow, 888p. We would like to acknowledge the NASA Glenn and 13. Krist, S.L., Biemn, RT., and Rmey, C.L., 1996, Langley Research Centers, especially Dr. Dennis M. CFL3D User's Manual (Version 5.0), NASA Langley Bushnell, Charles W.M cClinton, David W.L am, and Research Center, 3 1 1 p. Curt Snyder from the Naval Air System Team for 14. Gilinsky, M., Lebedev, M.G., and Yakubov, I.R., 19W, Modeling of Gas Flows with Shock waves, Publ.: interest and support to our research. This research "Maahinostroenie", 198p.,( in Russian). was partially conducted under the NASA grants: NAG-3-2249, 2422 and 2495, and under the supporting CRDF grant RE1-2068.W e would like to thank Dr. John M. Seiner and Dr. Jay C. Hardin for their attention, interest to our research, review and useful suggestions. M. REFERENCES 1. Seiner, J.M. and Gilinsky, M.M., 1997, Nozzle Thrust Optimization while Reducing Jet Noise, AIAA J, No. 3. 2. Seiner, J.M., and Gilinsky, M.M., 1995, Nozzle Thrust Optimization while Reducing Jet Noise, 26th AIAA Fluid Dynamics Conference, June 19-22,1995/SanD iego, CA. 3. Gilinsky, M., and Blankson, I.M., 2000, Internal Design Applications for Inlet and Nozzle Aeroperformance. Improvement, AIAA Paper #OO-3170, 36th 9 5*ol 1.8 7.3 2.0 3.5 1.1 J.3.