Search

A hybrid vortex filament scheme with the capability of simulating bounded turbulent flows is described. Viscous generation of new vortex elements at solid surfaces is accomplished through the intermediary step of solving the viscous vorticity transport equation on a thin boundary mesh via a finite difference and finite volume method. The transitional and turbulent boundary-layer flow past a wide, finite thickness, flat plate with rounded edges is computed with a view toward validating the methodology and gaining new insight into the structural aspects of transition. The predicted mean velocity and related statistics are well matched to experimental and numerical data. The representation of the flow through vortex filaments reveals that a distinction needs to be made between vortical structures, on one hand, and the rotational motion that they produce in the flowfield, on the other hand. In particular, hairpin-shaped regions that are commonly found by vortex identification schemes are found to not be structures in their own right, but rather the rotational signature of raised furrows in the surface vortex sheet. The latter overlie low-speed streaks and evolve to include ejected mushroom-shaped structures as well as spanwise vortices associated with roll-up. DOI: 10.2514/1.J050224

An experimental study was conducted to determine the effect of single dielectric barrier discharge plasma actuators on turbulent boundary-layer separation control. Two-component particle image velocimetry and laser Doppler velocimetry measurements showed the effect of plasma actuators on ambient air, a canonical zero-pressure gradient turbulent boundary layer, and a two-dimensional turbulent boundary-layer separation from a convex ramp section. Different actuator configurations and control strategies were implemented. Spanwise actuators were operated in both steady and unsteady modes. Plasma streamwise vortex generators were also used to enhance boundary-layer mixing. Flow visualization using a high-speed camera captured the temporal aspects of the separation control process and helped discern the physical mechanisms associated with each actuation strategy. The steady spanwise actuation produced a wall jet effect that augmented near-wall momentum and reattached the separated boundary layer. The streamwise oriented actuators also showed effective control authority by creating counter-rotating vortices within the boundary layer that promote mixing of high and low momentum fluid. DOI: 10.2514/1.J050009

Two-dimensional evolution equations are derived as applied to flows in the near-wall jet and the Blasius boundary layer on a flat plate on which a mechanism of inviscid--inviscid infraction controls the development of large-sized short-scaled disturbances. The first one is an extension of the Korteweg--de Vries equation. As distinct from the shallow-water wave motion underlying the Kadomtsev--Petviashvili equation, the fluid parameters are not assumed to depend only weakly on the direction transversal to the oncoming stream. The second dynamical system provides a two-dimensional analog of the Benjamin--Davis--Acrivos equation. Simple line-soliton solutions are presented in both cases. A generalized Hirota function allows a pair of crossed solitons to be obtained at some distance from a solid surface in the near-wall jet. An oblique periodic nonlinear wave train pointed out for the Blasius boundary layer comes in place of the Tollmien--Schlichting waves when their amplitude attains sufficiently large values. DOI: 10.2514/1.40761

To examine the potential of micro vortex generators for shock/boundary-layer interaction control, a detailed experimental and computational study in a supersonic boundary layer at M = 3.0 was undertaken. The experiments employed a flat-plate boundary layer with an impinging oblique shock with downstream total-pressure measurements. The moderate Reynolds number of 3800 allowed the computations to use monotone-integrated large eddy simulations. The monotone-integrated large eddy simulations predictions indicated that the shock changes the structure of the turbulent eddies and the primary vortices generated from the microramp. Furthermore, they generally reproduced the experimentally obtained mean velocity profiles, unlike similarly resolved Reynolds-averaged Navier-Stokes computations. The experiments and monotone-integrated large eddy simulations results indicate that the microramps, for which the height is h [approximately equal to] 0.5[delta], can significantly reduce boundary-layer thickness and improve downstream boundary-layer health as measured by the incompressible shape function H. Regions directly behind the ramp centerline tended to have increased boundary-layer thickness, indicating the significant three- dimensionality of the flowfield. Compared with baseline sizes, smaller microramps yielded improved total-pressure recovery. Moving the smaller ramps closer to the shock interaction also reduced the displacement thickness and the separated area. This effect is attributed to decreased wave drag and the closer proximity of the vortex pairs to the wall. DOI: 10.2514/1.41776

Numerical simulation of shock waves interacting with multimaterial interface is immensely challenging, particularly when the embedded interface is retained as a sharp entity. The challenge lies in accurately capturing and representing the interface dynamics and the wave patterns at the interface. In this regard, the ghost fluid method has been successfully used to capture the interface conditions for both fluid-fluid and solid-fluid interfaces. However, the ghost fluid method results in over/underheating errors when shocks impact interfaces, and hence must be supplemented with numerical corrective measures to mitigate these errors. Such corrections typically fail for strong shock applications. Variants and extensions of the ghost fluid method have been proposed to remedy its shortcomings with mixed success. In this paper, the performance of approaches based on the ghost fluid method, in the case of strong shocks impinging on immersed solid boundaries in compressible flows, is evaluated. It is found that (from the viewpoint of simplicity, robustness, and accuracy) a reflective boundary condition used in conjunction with a local Riemann solver at the interface proves to be a good choice. The method is found to be stable, accurate, and robust for wide range problems involving strong shocks interacting with embedded solid objects. DOI: 10.2514/1.43153

Numerical studies were carried out to investigate the effects of turbulent boundary layers on panel flutter at supersonic speeds. In this study, Reynolds-averaged Navier-Stokes equations were solved to take into account the turbulent boundary layer and its viscous effects. First, the fluid-structure coupling code was validated. The computed flutter boundaries agreed well with experimental data. Moreover, the results showed that the viscous effects were important and should be taken into account for flutter computation. Second, the boundary-layer effects were investigated in the Mach number range of 1.0-2.4. We compared the Reynolds-averaged Navier-Stokes computation with the inviscid computation and discussed the differences between them. We found that the boundary layer has not only a stabilizing effect but also a destabilizing effect, depending on the Mach number. The most important finding is that the flutter dynamic pressure slowly increases due to the boundary layer as the Mach number increases. In addition, the design boundary methodology was reviewed in terms of the turbulent boundary-layer effect, which will be helpful for the development of a new boundary-layer correction for the design boundary. DOI: 10.2514/1.35786

An Euler solver with boundary-layer option, overset/sheared Cartesian mesh, and automated mesh-generation capability has been developed for aeroelastic applications to complex aircraftlike configurations. The automated mesh-generation scheme call automatically generate a block mesh by extending the grid lines from surface meshes to the entire flowfield, in this way, the surface mesh can adopt the paneling definition required by commonly practiced aerodynamic panel methods as input. The overset/sheared-Cartesian-mesh capability allows convenient modeling of very complex configurations, such as an aircraft with external stores, in which the aircraft and stores are modeled by different blocks of meshes. This paper does not describe a novel computational method nor provide new insight into the characteristics of physical problem. Rather, it offers a user-friendly and cost-effective computational methodology that can be readily adopted by industry for rapid aeroelastic applications. DOI: 10.2514/1.43434

A new method for predicting the noise generated by a ducted rotor interacting with inhomogeneous and nonisotropic turbulence has been developed. The analytical formulation used a model of the two-point correlation function of the turbulent velocity in the space--time domain. The study focused on a specific condition where the dominant noise source was the interaction between a rotor and a casing turbulent boundary layer. The axial length scale of this turbulence was found to be large enough to generate unsteady lift that was correlated between multiple rotor blades. This led to tonal sound at the blade passing frequency in the absence of mean velocity variations. The analytical formulation was validated with a set of measurements obtained in a ducted rotor facility. The prediction accurately modeled both the tonal and broadband features of the sound spectra. DOI: 10.2514/1.43271

The interface condition provides flexibility to handle complex geometries on a set of structured grids. However, when used with a compact scheme to apply in computational aeroacoustics, its accuracy and stability may be degraded seriously by the boundary treatment that arises at a grid interface. To minimize this effect, a modified characteristic condition is incorporated in the finite difference compact scheme by using an explicit difference form at the boundary. The extension of this approach is naturally attained through the coordinate transformation of convective terms across an interface in multidimensions. The validity of current development is demonstrated through a simulation of trailing-edge noise generation from an airfoil at a low Mach number. DOI: 10.2514/1.42110

Laminar-turbulent transition remains a critical issue in a number of cases, including drag reduction, performance prediction of high-lift systems, improved accuracy in general computational fluid dynamics, and reduction of computation cycles for development of optimization tools. Transition delay remains one of the most promising technologies for reducing air transport energy consumption, through natural or hybrid laminar flow control. The use of linear stability theory, either local or nonlocal, remains rather demanding in terms of knowledge and user interaction. Hence, a demand exists for simplified, robust, and accurate transition prediction tools to be inserted into general flow solvers, of boundary-layer or Reynolds-averaged Navier--Stokes types. The problem can be solved by developing transition criteria or database methods. In this last case, characteristics of an actual flow are derived from known solutions of model flows. ONERA, the French Aerospace Laboratory, has long been involved in the development of such methods, and the present paper aims at providing a comprehensive view of the tools developed in the second category, applicable from low-speed two-dimensional to transonic three-dimensional flows, and even to three-dimensional supersonic flows. DOI: 10.2514/1.42990

As a first step toward predicting rough-wall boundary-layer noise, the sound radiated from a single hemispherical roughness element and a pair of roughness elements in a turbulent boundary layer at [Re.sub.[theta]] = 7500 is investigated. The roughness height is 3.6% of the boundary-layer thickness, or 95 wall units. The flowfield is obtained from large-eddy simulation, and the results are validated against experimental measurements. Acoustic calculations are performed based on the Curle--Powell integral solution to the Lighthill equation for an acoustically compact hemisphere. The sound radiation is dominated by unsteady drag dipoles and their images in the wall. It is found that the spanwise dipole, which has been overlooked in previous studies of roughness noise, is of larger or similar strength compared with the streamwise dipole, and the viscous contribution to the drag dipoles is negligible in comparison with the pressure contribution. Important flow features contributing to sound radiation are identified by examining the unsteady surface-pressure field and the surrounding flow structures. Pressure fluctuations are strongest on the upstream part of the hemispheric surface near the base due to impingement of incoming turbulent eddies and their interaction with horseshoe vortices. On the back surface of the hemisphere, pressure fluctuations are relatively weak, indicating that shear-layer separation and vortex shedding do not produce significant self-noise from the hemisphere. In the case with two hemispheres, the wake of the upstream hemisphere is found to significantly enhance sound radiation from the downstream hemisphere, particularly in the streamwise direction and at high frequencies. DOI: 10.2514/1.42343

An analytical model for the unswept normal shock wave/turbulent-boundary-layer interaction control using an upstream and downstream unimorph piezoelectric flap actuator has been proposed. The amount of flap deflection controls the bleed/suction rate through a plenum chamber. The cavity allows rapid thickening of the boundary layer approaching a normal shock wave, which splits into a series of weaker shocks forming a lambda shock foot, leading to a reduction in the wave drag. The analysis provides an understanding of the control influences produced in an experimental investigation of an unswept normal shock wave/turbulent-boundary-layer interaction at a Mach number of 1.5. It has also been validated by application to the normal shock wave/boundary-layer interaction control system using mesoflaps for aeroelastic transpiration described in previous transonic/supersonic shock wave/ boundary-layer interaction studies.

A two-dimensional Shack--Hartmann wave-front sensor is used to study aero-optic distortion in turbulent boundary layers at transonic and hypersonic speeds, with and without gas injection. The large-scale motions in the outer layer, of the order of the boundary-layer thickness in size, are shown to dominate the aero-optic distortion. Gas injection always reduced the Strehl ratio, with helium injection generally giving lower Strehl ratios than nitrogen injection. The large aperture approximation is shown to be accurate for a wide variety of aberrations regardless of Mach number and gas injection. A new scaling argument for the root-mean-square phase distortion is proposed that appears to collapse the data better than previous models. DOI: 10.2514/1.41453

Simulations of compressible boundary-layer flow at three different Reynolds numbers ([Re.sub.[delta]] = 5.59 x [10.sup.4], 1.78 x [10.sup.5], and 1.58 x [10.sup.6]) are performed using a hybrid large-eddy simulation/Reynolds-averaged Navier--Stokes method. Variations in the recycling/rescaling method, the higher order extension, the choice of primitive variables, the Reynolds-averaged Navier--Stokes to large eddy simulation transition parameters, and the mesh resolution are considered in order to assess the model. The results indicate that the present model can provide good predictions of the mean-flow properties, second-moment statistics, and structural features of the boundary layers considered. Normalized turbulent statistics in the outer layer are found to be independent of Reynolds number, similar to incompressible turbulent boundary layers. DOI: 10.2514/1.41598

This paper presents an experimental and computational study of a passive version of the air-jet vortex-generator flow control system for the delay or suppression of trailing-edge boundary-layer separation and subsequent stall. The results show that a passive vortex-generating jet flow control system, which uses the high pressure front the leading-edge attachment line to feed the upper-surface air jets, can effectively delay trailing-edge separation and subsequent stall to higher angles of attack, without the need for any active energy input and without significant drag penalty. DOI: 10.2514/1.41986

Direct numerical simulations are carried out on boundary-layer flow at Mach 6 over a porous surface, in which a Mack mode of instability is excited. The pores are resolved rather than modeled, allowing an evaluation to be made of the accuracy of simplified analytical models used in previous investigations based on linear stability theory. It is shown that the stabilizing effect of porosity is stronger in the simulations than in the corresponding theory for both two- and three-dimensional pores. From comparisons of spanwise grooves, streamwise slots, and square pores, it appears that the detailed surface structure is not as important as the overall porosity, and the hydraulic diameter is able to collapse the results for different pore shapes to good accuracy. When the porous surface consists of fewer larger pores, the flow is noisier, with sound waves generated at the pore edges. DOI: 10.2514/1.43388

Means of controlling the flow over a large spherical protuberance were examined. The role of suction around the base of the protuberance in reducing or even eliminating the necklace vortex created by the protuberance was considered. In the absence of suction, this vortex was lifted by the low base pressure existing behind the protuberance into the wake, thus affecting the turbulence level along its path. Large vortex generators placed upstream of the protuberance were able to delay local separation of the flow over the protuberance, thus affecting the symmetry of the wake and the level of turbulence on one side or the other. Observations made using flow visualization were supplemented by hot-wire measurements. The experiments were carried out at low speed at Reynolds numbers that did not exceed 3 x [10.sup.5].

Experiments were performed to investigate the flow in a Mach 2.9 shock wave turbulent boundary-layer interaction at a Reynolds number based on momentum thickness of 2400. The flow configuration was a nominally two-dimensional 24 deg compression ramp, which exhibited a separation bubble in the corner region. Mean flow quantities, including the velocity profile upstream and downstream of the corner, and the wall pressure through the interaction were measured. Filtered Rayleigh scattering was used to visualize the flow structure in the interaction and provide quantitative measurements of the turbulent structure angle and the intermittency of the boundary-layer edge turbulence. The shock motion was characterized by measuring the fluctuating wall pressure. The results indicate that, compared to previous measurements obtained at higher Reynolds numbers of 60-80,000, the separation bubble is approximately twice as long, the root mean square of the wall-pressure fluctuations has a relatively smaller peak, and the intermittency of the wall-pressure signal in the shock-foot region is attenuated. As in the high-Reynolds-number case, the shock motion has a broadband frequency distribution with a peak slightly below 1 kHz. The mean flow quantities, root mean square wall-pressure-fluctuation profile, wall-pressure signal, and shock-motion frequency agree well with the direct numerical simulation of a previous study at matching conditions.

The wall shear-stress sensor [MPS.sup.3] based on flexible micropillars has been used to experimentally assess the two- dimensional wall shear-stress distribution in turbulent duct flow at moderate Reynolds number. The sensor covers an area of 90 x 125 viscous length scales along the streamwise and spanwise directions, respectively. Applying Taylor's hypothesis allows a further increase in the streamwise extension. Preliminary results evidence the coexistence of low- and high-shear regions representing footprints of near-wall coherent structures. The low-shear regions resemble long meandering bands interrupted by local high-shear-stress regions. A qualitative comparison evidences the structures detected in the present study to be similar to wall shear-stress distributions reported in the literature and to structures found in higher regions of the boundary layer. These preliminary findings indicate the potential of the micropillar sensor concept to be capable of detecting the multidirectional planar wall shear-stress distribution in turbulent flows. A detailed discussion of the wall shear-stress characteristics and of the geometric properties of the streaklike structures is beyond the scope of this paper, and a more intrusive discussion needs to be done in the future.

This paper first discusses the constitutive relations and surface catalytic model for direct numerical simulation of wall-bounded turbulence including finite-rate chemistry and gas--surface interaction and then provides a systematic procedure to test the validity of the simulations by dividing the whole problem into different components and testing each component separately. Namely, comparisons against similarity solutions and other established hypersonic boundary-layer solutions are used to test the validity of laminar mean flow with and without gas-phase chemical reactions; comparisons against the analytic solution for the one-dimensional diffusion equation are used to test the validity of the surface catalysis boundary condition; and comparisons against empirical predictions, detailed experimental data and linear stability theory are used to test the validity of turbulent boundary-layer solutions. DOI: 10.2514/1.38318

This paper presents a review of data on controlling the boundary layer attachment by a plasma actuator with high-voltage pulsed-periodic nanosecond excitation. Actuator-induced gas velocities show near-zero values for nanosecond pulses. The measurements performed show overheating in the discharge region at fast ([tau] [equivalent] 1 [micro]s) thermalization of the plasma inputed energy. The mean values of such heating of the plasma layer can reach 70, 200, and even 400 K for 7-, 12-, and 50-ns pulse durations, respectively. The emerging shock wave together with the secondary vortex flows disturbs the main flow. The resulting pulsed-periodic disturbance causes an efficient transversal momentum transfer into the boundary layer and further flow attachment to the airfoil surface. Thus, for periodic pulsed nanosecond dielectric barrier discharge, the main mechanism of impact is the energy transfer to and heating of the near-surface gas layer. The following pulse-periodic vortex movement stimulates redistribution of the main flow momentum. The experiments performed here have shown high efficiency of the given mechanism to control boundary layer separation, lift, and drag force coefficients, and acoustic noise reduction in the Mach number range of 0.05 to 0.85. DOI: 10.2514/1.38113

Oil-film visualization patterns of the flow on a classical wing and on a wing with a wavy surface at various angles of attack are analyzed. For low Reynolds numbers (Re < 2 x [10.sup.5]), the boundary-layer structure on a wavy wing is demonstrated to differ significantly from that on a classical wing. As a result, leading-edge stall on a wavy wing occurs later than on a smooth wing. The present activities form the basis for a systematic study of the boundary-layer structure on a wavy wing. For this purpose, a series of investigations is performed. Boundary-layer parameters on classical and wavy-surface wings at the angle of attack [alpha] = 0 deg and the chord Reynolds number Re = 1.7 x [10.sup.5] are studied experimentally. The combined experiment includes oil-film visualization, measurements of pressure distributions, and flow velocity. On both wings, substantial differences are found in the transition position and in the transition scenarios realized on the wavy-surface wing along its streamwise grooves and humps. However, some similarity in the disturbance growth in both separation-bubble regions is found.

An experimental study has been conducted to investigate the effects of oscillation at a range of Strouhal numbers (0.0006 to 0.0113) on the boundary-layer development of a flat plate under the influence of favorable pressure gradients. The Reynolds number was [10.sup.6] based on the flat-plate chord length. An inverted airfoil placed in the turbulent region over the flat plate was used to create the effects of favorable pressure gradients. Hot-wire anemometry was used to measure the velocities in the streamwise direction. The flow separates near the rounded leading edge, causing the formation of the separation bubble. It is found that the time-averaged velocity and turbulence intensity profiles remain mostly unaffected by oscillation in the turbulent flow regions (II and III). The boundary-layer properties such as the boundary-layer displacement thickness, momentum thickness, and the shape factor remain almost constant. However, in region I, in which the flow is in its transitional state, oscillation slows the reattachment process under the influence of favorable pressure gradients.

An experimental study has been conducted to investigate the effects of oscillation, of varying amplitude (0 to 0.03 m) and frequency (0 to 6 Hz), on the boundary-layer development of a flat plate under the influence of adverse pressure gradients. The Reynolds number was [10.sup.6] based on the flat plate chord length. The study employed a subsonic wind tunnel facility at a freestream velocity of 16 m/s. A cylinder placed at various locations over the flat plate within the boundary layer was used to create the effects of adverse pressure gradients. Surface pressures were measured along the centerline of the flat plate to study the flow quantitatively. Oil flow visualizations were also obtained to examine the flow qualitatively. Boundary-layer surveys were conducted to examine the effect of oscillation on the mean velocity, turbulence intensity, and boundary-layer thickness profiles.

We investigate the performance of two recently developed subgrid-scale models, the approximate deconvolution model and the high-pass filtered Smagorinsky model, in large-eddy simulations of laminar-turbulent transition in a supersonic boundary layer. Subharmonic transition in a boundary layer at a freestream Mach number of 4.5 and a Reynolds number (based on initial displacement thickness) of 10,000 is considered, which has been studied previously in detail by direct numerical simulations. For computational efficiency, the temporal simulation approach has been adopted. The discretization is based on Fourier collocation and various high-order finite difference schemes in the wall-parallel and wall-normal directions, respectively. Large-eddy simulations results are assessed by comparing statistical and instantaneous quantities during transition with data obtained from a sufficiently resolved direct numerical simulation. The results show that the large-eddy simulations accurately reproduce the direct numerical simulations data from the slightly disturbed laminar flow through transition into the turbulent stage, with a computational effort of two orders of magnitude less than the direct numerical simulations. Both subgrid-scale models are formulated locally in space and in a fully three-dimensional manner and do not need an ad hoc adaptation to nonturbulent or near-wall regions. DOI: 10.2514/1.15048

A robust scheme for generating unstructured viscous meshes, using elliptic partial differential smoothing, is presented. Linear-elasticity relationships, borrowed from the structural mechanics discipline, provide the governing partial differential equations for smoothly displacing the existing mesh, making room for new viscous layers. These viscous layers are allowed to advance up adjacent, nonviscous boundaries. A process of inserting layers on alternating boundary groupings enables the creation of high-quality hexahedral elements in juncture regions. Viscous layer height is allowed to vary, depending on the size of the local inviscid mesh. Three-dimensional results are included to illustrate the use of the method to produce high-quality viscous meshes.

A new rational theory of incompressible turbulent boundary-layer flows having a large velocity defect is presented on basis of the Reynolds-averaged Navier-Stokes equations in the limit of infinite Reynolds number. This wake-type formulation allows for, among others, the prediction of singular solutions of the boundary-layer equations under the action of a suitably controlled adverse pressure gradient, which are associated with the onset of marginally separated flows. Increasing the pressure gradient locally then transforms the marginal-separation singularity into a weak Goldstein-type singularity occurring in the slip velocity at the base of the outer wake layer. Interestingly, this behavior is seen to be closely related to (but differing in detail from) the counterpart of laminar marginal separation, in which the skin friction replaces the surface slip velocity. Most important, adopting the concept of locally interacting boundary layers results in a closure-free and uniformly valid asymptotic description of boundary layers that exhibit small, closed reverse-flow regimes. Numerical solutions of the underlying triple-deck problem are discussed.

A pilot study was conducted on a flapped semispan model to investigate the concept and viability of near-wake vortex management by means of boundary layer separation control. Passive control was achieved using a simple fairing and active control was achieved via zero mass-flux blowing slots. Vortex sheet strength, estimated by integrating surface pressures, was used to predict vortex characteristics based on inviscid roll-up relations and vortices trailing the flaps were mapped using a seven-hole probe. Separation control was found to have a marked effect on vortex location, strength, tangential velocity, axial velocity, and size over a wide range of angles of attack and control conditions. In general, the vortex trends were well predicted by the inviscid roll-up relations. Manipulation of the separated flow near the flap edges exerted significant control over either outboard or inboard edge vortices while producing small lift and moment excursions. In summary, separation control has the potential for application to time-independent or time-dependent wake alleviation schemes, where the latter can be deployed to minimize adverse effects on ride-quality and dynamic structural loading. DOI: 10.2514/1.19664

In this paper, we formulate the stacking sequence design of laminated composite flat panels for maximum supersonic flutter speed and maximum thermal buckling capacity. The design is constrained to have stable limit cycle oscillation in the vicinity of the flutter boundary. For the purpose of comparison, we consider both linear and nonlinear panel flutter analyses. Because of manufacturing constraints, the stacking sequence of the laminate is selected from a discrete set of orientation angles. The plate is modeled using von Karman plate equations and assumed to be subjected to uniform temperature differential. The aerodynamic load on the plate is computed using third-order piston theory. The plate equations of motion are derived using Hamilton's principle and discretized using the Rayleigh-Ritz method. The nonlinear panel flutter analysis is performed using an approximate perturbation approach. The Pareto fronts describing the trade-off between thermal buckling and flutter are generated for different panel aspect ratios. It is observed that optimal designs obtained for both linear and nonlinear flutter analyses are similar, with only the predicted flutter speeds different. This suggests that, for the purpose of preliminary stacking sequence design, linear flutter analysis is sufficient. DOI: 10.2514/1.19532

An optimal thermal protection system design for a spacecraft operating in extreme environments of thermal and acoustic loading is of significant importance for today's military space missions. These military space missions tend to push the envelope of the spacecraft's capabilities to the extreme, requiring robust performance of all key systems to assure a successful mission. Because thermal protection system protects the entire spacecraft, its survival from these extreme conditions is critical to the safety of the mission. The design criteria for the thermal and acoustic loading conditions tend to be conflicting. To reduce thermal stress, free boundaries that allow for expansion of the thermal protection system are desirable. However, the random sound pressure level fluctuations associated with, for example, engine noise during ascent, subject the thermal protection system to a wide-band random excitation. Therefore, if the thermal protection system possesses low frequency modes, damaging strains could result. To limit the magnitude of the strains, higher frequency (stiffer) designs with fixity at the boundary are desirable. Therefore, when designing the thermal protection system within these two simultaneous operating environments, the safe operating region is often severely restricted. This research aims at evolving a design that satisfies these two design requirements. DOI: 10.2514/1.15906

Detailed experiments were made to gain an improved understanding of relaminarization on swept wings under high-lift conditions. Two swept wings with different airfoil sections and flaps were tested, and measurements were made consisting of surface-pressure distributions and wall shear-stress fluctuations in the leading-edge zone. Measurements were made at a chord Reynolds number of 1.3 x [10.sup.6] with wing-sweep angles of 45 and 60 deg and model incidence varied in the range of 3 to 18 deg in discrete steps. Following the attachment-line transition, the turbulent boundary layer relaminarized under certain conditions due to the acceleration around the wing's leading edge. Broad features of relaminarization and subsequent retransition are discussed based on wall shear-stress-fluctuation data. The present dataset shows that relaminarization is likely if the maximum value of the acceleration parameter [K.sub.s] (evaluated along the external inviscid streamline) is greater than about 3 x [10.sup.-6], in agreement with earlier findings. Much higher maximum values of [K.sub.s] result in significant reduction of intermittency during relaminarization. After relaminarization, the boundary layer retransitions across a separation bubble. Characteristics of the postrelaminarization separation bubbles are shown to be similar to those of bubbles occurring in more pristine laminar boundary layers. DOI: 10.2514/1.20035

Major challenges to low speed microflight are the transient and time-averaged velocities arising from the atmospheric boundary layer, particularly turbulence a few meters above the ground. In this paper, prior work on the temporal and spatial characteristics of the atmospheric boundary layer close to the ground, and the relative turbulence as perceived by a moving craft, are considered. New measurements are described that document the time-averaged and transient velocities at a height of 4 m above the ground. These were made using a bank of four multihole pressure probes laterally separated by 150 and 50 mm on a mast above a test car. Transient flow pitch angles were investigated and it was found, that the overall variation with lateral separation decreased relatively slowly with reducing separation, but that both this and the pitch angle coherence may be described nondimensionally. As the slow decrease in pitch variation with lateral spacing implies that the roll inputs arising from vertical fluctuations would increase with reducing span, it is speculated that increasingly active and authoritative control systems are required. DOI: 10.2514/1.22670

Over the past decade almost every aspect of our traditional beliefs about wall-bounded flows has been challenged. Some beliefs have been abandoned and others modified as new theories and new measurements have abounded. No longer are direct measurements summarily dismissed because they contradict a long-standing theory, and even those experimentalists who still believe in the log law only apply it outside [y.sup.+] of several hundred. In this meeting I have been invited to look back at some of the early contributions in this revolution. I shall review briefly a few of the more controversial ideas, and highlight some of the reasons that led us in a different direction. In addition to pointing out some successes, I will discuss some failures and open questions. I will even offer a few new ideas, and of course renew some old challenges. DOI: 10.2514/1.19951

Recent efforts by the present authors have focused on the fundamental multiscaling behaviors of the time averaged dynamical equations of wall turbulence. These efforts have generated a number of new results relating to dynamical structure, as well as a new mathematical foundation. Central to this has been the development of the so-called method of scaling patches. This method provides a formalism for determining scaling behaviors directly from the indeterminate equations. A general description of this methodology is provided herein, and in doing so its connections to well-established scaling notions are identified. Example problems for which the method has been successfully applied includes turbulent boundary layer, pipe and channel flows, turbulent Couette-Poiseuille flow, fully developed turbulent heat transfer in a channel, and favorable pressure gradient boundary layers. DOI: 10.2514/1.18911

This paper presents a consistent theory of self-similarity and of equilibrium in the outer region of turbulent boundary layers that explains recent experimental findings on the subject, including new ones presented here. The theory is first presented in a general form where the outer scales are left unspecified and it is not assumed that the mean velocity defect and the Reynolds stresses share a common velocity scale. It is shown that the main results of the traditional similarity theory remain valid even in this case. Common outer scaling with the Zagarola-Smits length and velocity scales is then chosen. A new pressure gradient parameter is introduced to characterize the local effect of the pressure gradient in all flow conditions including strong adverse-pressure-gradient conditions. By analyzing several adverse-pressure-gradient flow cases, it is shown that self-similarity of the mean velocity defect profile is reached in all cases in localized but significant flow regions. The same is, however, not true of the Reynolds stress profiles. In agreement with the similarity analysis, the self-similar velocity defect profile is found to be a function of the pressure gradient and most flows studied here are only in an approximate state of equilibrium in the region of self-similar defect profiles despite the excellent collapse of the profiles. DOI: 10.2514/1.19234

An experimental study of supersonic inlet buzz is presented. This study was carried out on a mixed compression rectangular inlet model; tests were done at Mach numbers ranging from 1.8 to 3, with and without bleed. Inlet flows were analyzed thanks to Schlieren videos and signal processing of unsteady pressure recordings. Two kinds of buzz were observed: 'little buzz,' which is thought to correspond to an acoustic resonance phenomenon excited by the presence of a shear layer under the cowl lip (Ferri criterion), and 'big buzz,' which seems to be triggered by a boundary layer separation on the compression ramps (Dailey criterion). In some cases, little buzz could be virtually suppressed by the introduction of a bleed. The frequency of big buzz is shown to be already present in the flow before the onset of large-amplitude oscillations, which suggests that the underlying mechanism of big buzz, probably linked to acoustics, already exists before buzz starts.

A large eddy simulation was made of transonic flow over a two-dimensional bump where shock wave turbulent boundary layer interaction takes place. Grid refinement and the effect of the domain width were investigated. Special care was taken to ensure physically correct inlet boundary conditions. The shock wave turbulent boundary layer interaction induces strong separation of the boundary layer and events such as bursting events in the incoming boundary layer, and creation of large flow scale structures behind the shock are detected. However, the shock features no large scale movement even though, according to several sources, the prerequisites for such movement are fulfilled.

This paper presents detailed investigations related to active transition control in laminar separation bubbles. The investigations rely on direct numerical simulations based on the complete Navier--Stokes equations for a flat-plate boundary layer. A laminar separation bubble is created by imposing a streamwise adverse pressure gradient at the freestream boundary of the integration domain. Different steady and unsteady boundary layer disturbances are then introduced at a disturbance strip upstream of separation and their effects on the separation bubble are studied. It is shown that the size of the separated region can be controlled most efficiently by very small periodic oscillations, which lead to traveling instability waves that grow to large levels by the hydrodynamic instability of the flow. Indications for the preferred frequency of these waves can be obtained from linear stability theory, but since the problem is nonlinear, only direct numerical simulations can really qualify or disqualify the predictions. Overall, it turns out that unsteady two- or three-dimensional disturbances have a stronger impact on the size of the bubble than steady disturbances, because they directly provide initial amplitudes for the laminar-turbulent transition mechanism.

Results of computations of aerodynamic flows with attached and separated boundary layers with adverse pressure gradients are presented. The objective is to compare the predictive accuracy of 11 eddy-viscosity turbulence models frequently employed for computing aeronautical flows. The turbulence models considered are the algebraic model of Baldwin and Lomax; the one-equation model of Spalart and Allmaras along with two additional variants of it; six different k, [omega] models; and the nonlinear eddy-viscosity model of Wallin and Johansson. Starting with a boundary-layer flow with a nominally zero pressure gradient, flows with a successively increased adverse pressure gradient are investigated. Computational results are compared with each other and to experimental data. It is found that the agreement between computational and experimental results varies between the turbulence models also for the simpler flow cases. Predictive accuracy obtained with a given model for a given flow case varies between the different flow variables evaluated. The 'best' turbulence model changes from flow case to flow case. Models using transport equations for the eddy viscosity (or a related quantity) typically show better response to a pressure gradient than the algebraic Baldwin-Lomax model. However, no general conclusion on which model to use for adverse-pressure-gradient flows is drawn.

Direct aeroacoustic computations require nonreflective boundary conditions that allow disturbances to leave the domain freely without anomalous reflections. In the present paper we analyze a series of nonreflective boundary conditions already published in the literature and propose an improved outflow nonreflective boundary condition with reflection characteristic that is reduced greatly. For the solution of the linearized Euler equations, a sixth-order compact finite-difference algorithm is used together with a sixth-order explicit digital filter that suppresses the high-frequency spurious oscillations in the solution. A number of representative test cases are presented. The new outflow boundary condition is recommended for the simulation of sound produced by turbulence.

A fundamental study is made of a rotating boundary layer (BL) in the inboard region of a tilt-rotor blade, using an embedded laser Doppler velocimetry probe inside a rotating blade. Velocity measurements have been performed to get insight into the BL behavior around a blade section close to the rotor hub. BL transition and separation phenomena are investigated on the basis of both chordwise and spanwise velocity profiles measured for different collective pitch angles on a two-bladed hovering rotor. The velocity experimental data bank appears to be in good agreement with theoretical velocity profiles in attached flow conditions (laminar, transitional, or turbulent). Moreover, validation of a transitional criterion based on the second thickness of energy [[delta].sub.3] of the BL is extended to the rotational flow configuration. Usually such a configuration classically exhibits a separation of the BL for high collective pitch angles. However, in the particular case of a high twisted tilt-rotor blade operating in hover, centrifugal effects are shown to be responsible for an important delay in the BL separation process.

A simple average of local enthalpy over the velocity profile is proposed as the proper definition of reference enthalpy for the purpose of quasi-one-dimensional treatment of compressible boundary layers. Use of Van Driest's nearly exact solutions of the laminar boundary-layer equations shows that this definition produces Eckert's reference enthalpy formulation for the special case of an adiabatic wall. For surfaces other than adiabatic, either Eckert's form should be replaced by that of Young and Janssen, or the coefficient in Eckert's viscous heating term should be slightly modified. A similar analysis was conducted for turbulent flows using Whitfield and High's simplified solutions of the turbulent boundary-layer equations. Dorrance's derivation of reference quantities is also addressed. This work provides a theoretical basis for the empirical reference enthalpy formulas of Eckert and others and supplies practical expressions for the reference enthalpy of both laminar and turbulent compressible boundary layers.

The influence of two-dimensional humps and steps on the stability characteristics of a two-dimensional laminar boundary layer is investigated by means of direct numerical simulations (DNS). The localized surface irregularity is hereby modeled within the Cartesian grid by assigning body forces over surfaces that need not coincide with grid lines. Compared to the use of a body-fitted coordinate system, this method saves memory and computation time. The method is validated by grid refinement tests as well as by a comparison with water channel experiments. The DNS results for the steady flow over a rectangular hump, as well as for an instability wave traveling over a hump, show good agreement with the experimental ones. Simulation results show that a localized hump destabilizes the laminar boundary layer, whereas a forward facing step stabilizes it. The destabilization is stronger when the height or the width of the localized hump are increased. A rounded shape of the hump is less destabilizing than a rectangular shape with sharp corners. The parameter that shows the strongest influence on the stability characteristics of the boundary layer is clearly the height of the localized hump.

A model experiment has been designed in the frame of a coordinated European project called AEROMEMS. This model is a two-dimensional bump in a boundary-layer wind tunnel, which mimics the adverse pressure gradient on the suction side of an airfoil at the verge of separation. The flow is characterized mainly by hot-wire anemometry. The integral parameters and the mean velocity profiles are compared to existing models and to Navier-Stokes computations. A new scaling and a new mixing length model are proposed and validated on the Ludwig and Tillman data also.

Aeroelastic mesoflaps for recirculating transpiration have been investigated in an effort to control shock/boundary-layer interactions (SBLIs) through passive cavity recirculation. The mesoflap concept utilizes a matrix of small flaps covering an enclosed cavity that are designed to undergo local aeroelastic deflection to achieve proper mass bleed or injection when subjected to gasdynamic pressure loading. Experiments were performed to investigate the applicability of the mesoflap concept for oblique shock interaction by employing shadowgraph flow visualizations, surface pressure measurements, and mean and fluctuating velocity measurements, along the spanwise midplane of the shock intersection. The experiments were conducted in a Mach 2.41 supersonic wind tunnel operating at a unit Reynolds number of 57 x [10.sup.6] [m.sup.-1]. With the thickest mesoflap arrays in place, the leading shock formed at the location of the first flap and the boundary-layer thickness at shock impingement was greater due to flow injection through the upstream flaps. However, the thinnest mesoflap arrays yielded a somewhat reduced boundary-layer thickness downstream of the interaction as a result of the tangential bleeding by the last flaps. Stagnation pressure profiles for the thinnest arrays also showed improved recovery downstream of the SBLI as compared to the solid-wall case. However, further study is needed to investigate three-dimensional effects and to determine whether this control strategy provides significant performance improvements for flow conditions more consistent with actual inlets.

The shapes (flame-sheet and luminous-flame boundaries) of steady weakly buoyant round hydrocarbon-fueled laminar-jet diffusion flames in still and coflowing air were studied both experimentally and theoretically. Flamesheet shapes were measured from photographs using a CH optical filter to distinguish flame-sheet boundaries in the presence of blue C[O.sub.2] and OH emissions and yellow continuum radiation from soot. Present experimental conditions included acetylene-, methane-, propane-, and ethylene-fueled flames having initial reactant temperatures of 300 K, ambient pressures of 4-50 kPa, jet-exit Reynolds numbers of 3-54, initial air/fuel velocity ratios of 0-9, and luminous flame lengths of 5-55 mm; earlier measurements for propylene- and 1,3-butadiene-fueled flames for similar conditions were considered as well. Nonbuoyant flames in still air were observed at microgravity conditions; essentially nonbuoyant flames in coflowing air were observed at small pressures to control effects of buoyancy. Predictions of luminous flame boundaries from soot luminosity were limited to laminar smoke-point conditions, whereas predictions of flame-sheet boundaries ranged from soot-free to smoke-point conditions. Flame-shape predictions were based on simplified analyses using the boundary-layer approximations along with empirical parameters to distinguish flame-sheet and luminous-flame (at the laminar smoke point) boundaries. The comparison between measurements and predictions was remarkably good and showed that both flame-sheet and luminous-flame lengths are primarily controlled by fuel flow rates with lengths in coflowing air approaching 2/3 of the lengths in still air as coflowing air velocities are increased. Finally, luminous flame lengths at laminar smoke-point conditions were roughly twice as long as flame-sheet lengths at comparable conditions because of the presence of luminous soot particles in the fuel-lean region of the flames.

Particle image velocimetry and high-frequency response wall pressure measurements have been used to investigate the relationship between upstream turbulent boundary-layer properties and the unsteady separation shock behavior in a Mach 5 unswept compression ramp interaction. No correlation is found between variations in the incoming boundary-layer thickness and the separation shock foot position, as has been suggested in earlier work. However, the mean velocity profile, conditioned on the separation shock foot position, exhibits a subtly fuller shape when the shock is downstream than when it is upstream. More significantly, a clear correlation is observed between positive streamwise velocity fluctuations in the lower third of the upstream boundary layer and downstream shock motions, and vice versa. The strongest correlations are found for velocity fluctuations with frequencies of about 4-10 kHz, which is significantly lower than the frequencies that characterize the large-scale structures in the boundary layer (40 kHz), although spatial limitations in the transducer array may limit the instrument sensitivity to this lower range. These results are qualitatively consistent with the simple physical principle that a fuller velocity profile imparts increased resistance to separation to the boundary layer and, hence, causes downstream shock motion, whereas a less-full velocity profile is associated with lower resistance to separation and, hence, upstream shock motion.

An anisotropic viscous Cartesian grid method based on a [2.sup.N] tree data structure is developed. The method is capable of handling complex geometries automatically. In addition, viscous boundary layers can be computed with high resolution, using automatically projected high aspect ratio viscous layer grids. Compared with a widely used Octree data structure, the [2.sup.N] tree data structure supports anisotropic grid adaptations in any of the coordinate directions. Therefore, key flow features such as shock waves, wakes, and vortices can be captured in a very efficient manner. To handle the adaptive viscous Cartesian grid, an implicit, second-order, finite volume flow solver supporting arbitrary grids has been developed. A linearity-preserving least-squares solution reconstruction algorithm is used to achieve second-order accuracy. Furthermore, several directional adaptation criteria are developed and tested. The overall grid generation, flow simulation, and grid adaptation methodology is then demonstrated for a variety of flow problems, including a case of supersonic turbulent flow over a high-angle-of-attack missile configuration.

Experiments to investigate the transition process in hypervelocity boundary layers were performed in the T4 free-piston shock tunnel. An array of thin-film heat-transfer gauges was used to detect the location and extent of the transitional region on a 1500 mm long x 120 mm wide flat plate, which formed one of the walls of a duct. The experiments were performed in a Mach 6 flow of air with 6- and 12-MJ/kg nozzle-supply enthalpies at unit Reynolds numbers ranging from 1.6 x [10.sup.6] to 4.9 x [10.sup.6] [m.sup.-1]. The results show that the characteristics typical of transition taking place through the initiation, growth, and merger of turbulent spots are evident in the heat-transfer signals. A 2-mm-high excrescence located 440 mm from the leading edge was found to be capable of generating a turbulent wedge within an otherwise laminar boundary layer at a unit Reynolds number of 2.6 x [10.sup.6] [m.sup.-1] at the 6-MJ/kg condition. A tripping strip, located 100 mm from the leading edge and consisting of a line 37 teeth of 2 mm height equally spaced and spanning the test surface, was also found to be capable of advancing the transition location at the same condition and at the higher enthalpy condition.

The effect of surface roughness and step changes of surface roughness on an unseparated incipient shockwave/turbulent boundary-layer interaction has been studied using complementary experimental and theoretical investigations. It was found that step changes in surface roughness lead to nonequilibrium boundary layers displaying different characteristics in the near-wall and far-wall regions. Both theory and experiment confirmed that the dominant effect of roughness, even in configurations featuring step changes of surface properties, is a change of boundary-layer displacement thickness and that the upstream influence of the interaction scales very well with the roughness modified incoming displacement thickness in all cases. The theory also identified small additional roughness effects, particularly for configurations featuring a smooth/rough wall transition ahead of the interaction.