Showing papers on "Flow separation published in 1989"

Airfoil self-noise and prediction

Abstract: A prediction method is developed for the self-generated noise of an airfoil blade encountering smooth flow. The prediction methods for the individual self-noise mechanisms are semiempirical and are based on previous theoretical studies and data obtained from tests of two- and three-dimensional airfoil blade sections. The self-noise mechanisms are due to specific boundary-layer phenomena, that is, the boundary-layer turbulence passing the trailing edge, separated-boundary-layer and stalled flow over an airfoil, vortex shedding due to laminar boundary layer instabilities, vortex shedding from blunt trailing edges, and the turbulent vortex flow existing near the tip of lifting blades. The predictions are compared successfully with published data from three self-noise studies of different airfoil shapes. An application of the prediction method is reported for a large scale-model helicopter rotor, and the predictions compared well with experimental broadband noise measurements. A computer code of the method is given.

Turbulent Boundary-Layer Separation

Abstract: This article summarizes our present understanding of the physical behavior of two-dimensional turbulent separated flows, which occur due to adverse pressure gradients around streamlined and bluff bodies. The physical behavior of turbulence is flow dependent, so detailed experimental infor­ mation is needed for understanding such flows and modeling their physics for calculation methods. An earlier review (Simpson 1 985) discussed in much detail prior experimental and computational work, and this was followed by an updated review of calculation methods only (Simpson 1 987). Here additional recent references are added to those cited in the two other works. By separation, we mean the entire process of departure or breakaway, or the breakdown of boundary-layer flow. An abrupt thickening of the rotational-flow region next to a wall and significant values of the normal­ to-wall velocity component must accompany breakaway, or otherwise this region would not have any significant interaction with the free-stream flow. This unwanted interaction causes a reduction in the performance of the flow device of interest (e.g. a loss of lift on an airfoil or a loss of pressure rise in a diffuser). It is too narrow a view to use vanishing surface shearing stress or flow reversal as the criterion for separation. Only in steady two-dimensional flow do these conditions usually accompany separation. In unsteady two­ dimensional flow the surface shear stress can change sign with flow reversal without the occurrence of breakaway_ Conversely, the breakdown of the boundary-layer concept can occur before any flow reversal is encountered. In three-dimensional flow the rotational layer can depart without the

The turbulent flow field around a circular cylinder

Abstract: The flow field around a circular cylinder mounted vertically on a flat bottom has been investigated experimentally. This type of flow occurs in several technical applications, e.g. local scouring around bridge piers. Hydrogen bubble flow visualization was carried out for Reynolds numbers ranging from 6,600 to 65,000. The main flow characteristic upstream of the cylinder is a system of horse-shoe vortices which are shed quasi-periodically. The number of vortices depends on Reynolds number. The vortex system was found to be independent of the vortices that are shed in the wake of the cylinder. The topology of the separated flow contains several separation and attachment lines which are Reynolds number dependent. In the wake region different flow patterns exist for each constant Reynolds number.

A natural low-frequency oscillation of the flow over an airfoil near stalling conditions

Abstract: An experimental and computational study of the low-frequency oscillation observed in the flow over an airfoil at the onset of static stall is presented. Wind-tunnel results obtained with two-dimensional airfoil models show that this phenomena takes place only with a transitional state of the separating boundary layer. It is noted that the flowfield does not involve a Karman vortex street. The experimental results agree well with the results of a two-dimensional Navier-Stokes code. The present study demonstrates that the low-frequency oscillations produce intense flow fluctuations which impart much larger unsteady forces to the airfoil than experienced by bluff-body shedding and which may represent the primary aerodynamics of stall flutter of blades and wings.

Direct numerical study of leading-edge contamination

Abstract: Instability, turbulence, and relaminarization in the attachment-line region of swept and unswept cylindrical bodies are studied by numerical solution of the full Navier-Stokes equations. The flow is simulated over a strip containing the attachment-line and treated as homogeneous in the spanwise direction; the disturbances decay exponentially upstream. Transpiration through the wall may be prescribed. The new method, which admits completely general disturbance, agrees with published linear-stability results, which were limited to an apparently restrictive form of disturbance. Fully developed turbulent solutions with sweep are generated and compare well with the experiment. The turbulence is subcritical (except for blowing), resulting in large hysteresis loops. By lowering the sweep Reynolds number, or increasing the suction, the turbulent flow is made to relaminarize. The relaminarization Reynolds number is much less sensitive to suction than the linear-stability Reynolds number. Extensive attempts to detect the postulated nonlinear instability of the unswept flow failed, suggesting that this flow is linearly and nonlinearly stable.

Direct simulation of a turbulent oscillating boundary layer

Abstract: The turbulent boundary layer under a freestream velocity that varies sinusoidally in time around a zero mean is considered. The flow has a rich variety of behaviors including strong pressure gradients, inflection points in the velocity profile, and reversal of the shear stress. A theory for the velocity- and stress profiles at high Reynolds number is formulated. Well-resolved direct Navier-Stokes simulations are conducted over a narrow range of Reynolds numbers. The flow is also computed over a wider range of Reynolds numbers using a new algebraic turbulence model. The results produced by the three approaches and by experiments are compared. Detailed phase-averaged statistical results from the direct simulations are provided to assist turbulence-model development.

Morphological adaptation of shape to flow: Microcurrents around lotic macroinvertebrates with known Reynolds numbers at quasi-natural flow conditions.

Abstract: Using Laser Doppler Anemometry we measured current velocities in the median plane around dead lotic macroinvertebrates in a flume which reproduced natural near bottom hydraulics. We investigated specimens of the gastropods Ancylus, Acroloxus, and Potamopyrgus, the amphipod Gammarus, and the larval caddisflies Anabolia, Micrasema, and Silo of various size, various alignment to the flow or which were otherwise manipulated in order to clarify certain questions of adaptation of shape or case building style to flow, or the effects of flow on field distribution patterns. The steepest velocity gradients close to the animals were found near areas of their bodies protruding furthest into the flow. In such regions the rates of potential diffusive exchange processes, the potential corrasion (abrasion through suspended solids), and, for larger specimens, the lift forces (directed towards the water surface) must be highest. Posterior of these areas growing boundary layers formed above those species whose upper contour was approximately parallel to the upstream-downstream direction of the flow. All specimens removed momentum from the flow and thus experience a drag force (directed downstream). From the complete data set we derived the following general conclusions about the physical effects of potential morphological adaptations, taking into consideration diffusion through boundary layers, corrasion, lift forces, friction and pressure drag forces: The physical significance of these five factors generally depends on the Reynolds number of an animal and is largely affected by flow separation, which was significantly related to the ratio of body length to height and the slope of the posterior contour. A simultaneous effective morphological adaptation to all five factors is physically impossible and, in addition, would have to change from life at low (e.g. a young, small specimen of a species) to life at high (e.g. a fully grown specimen of the same species) Reynolds number.

High-resolution upwind scheme for vortical-flow simulations

Abstract: One of the high-resolution upwind schemes called "MUSCL with Roe's average" is applied to vortical flow simulations. Two examples are considered. One is the leading-edge, separation-vortex flow over a strake-delta wing. The other is a high angle-of-attack supersonic flow over a space-plane-like configuration. The comparison with the central-difference solutions indicates that the present upwind scheme is less dissipative and thus has better resolution for the vortical flows. Thus, it is concluded that the use of proper unwind schemes is recommended for vortical flow simulations at a high Reynolds number, and verification of computed result is especially important for vortical-flow simulations.

On the Lagrangian description of unsteady boundary layer separation. Part 1: General theory

Abstract: Although unsteady, high-Reynolds number, laminar boundary layers have conventionally been studied in terms of Eulerian coordinates, a Lagrangian approach may have significant analytical and computational advantages. In Lagrangian coordinates the classical boundary layer equations decouple into a momentum equation for the motion parallel to the boundary, and a hyperbolic continuity equation (essentially a conserved Jacobian) for the motion normal to the boundary. The momentum equations, plus the energy equation if the flow is compressible, can be solved independently of the continuity equation. Unsteady separation occurs when the continuity equation becomes singular as a result of touching characteristics, the condition for which can be expressed in terms of the solution of the momentum equations. The solutions to the momentum and energy equations remain regular. Asymptotic structures for a number of unsteady 3-D separating flows follow and depend on the symmetry properties of the flow. In the absence of any symmetry, the singularity structure just prior to separation is found to be quasi 2-D with a displacement thickness in the form of a crescent shaped ridge. Physically the singularities can be understood in terms of the behavior of a fluid element inside the boundary layer which contracts in a direction parallel to the boundary and expands normal to it, thus forcing the fluid above it to be ejected from the boundary layer.

Turbulent flow separation control through passive techniques

Abstract: Several passive separation control techniques for controlling moderate two-dimensional turbulent flow separation over a backward-facing ramp are studied. Small transverse and swept grooves, passive porous surfaces, large longitudinal grooves, and vortex generators were among the techniques used. It was found that, unlike the transverse and longitudinal grooves of an equivalent size, the 45-deg swept-groove configurations tested tended to enhance separation.

Turbulent free shear layer mixing

Abstract: Some experimental data on turbulent shear layer growth, mixing and chemical reactions in the limit of fast kinetics are reviewed. The dependence of these phenomena on such fluid and flow parameters as Schmidt number, Reynolds number, and Mach number are discussed with the aid of some recent models and consequences deducible from the large scale organization of the flow.

A review of quasi-coherent structures in a numerically simulated turbulent boundary layer

Abstract: Preliminary results of a comprehensive study of the structural aspects of a numerically simulated number turbulent boundary layer are presented. A direct Navier-Stokes simulation of a flat-plate, zero pressure gradient boundary layer at Re0 = 670 was used. Most of the known nonrandom, coherent features of turbulent boundary layers are confirmed in the simulation, and several new aspects of their spatial character are reported. The spatial relationships between many of the various structures are described, forming the basis for a more complete kinematical picture of boundary layer physics than has been previously known. In particular, the importance of vortex structures of various forms to the generation of Reynolds shear stress is investigated.

Separation control on an airfoil by periodic forcing

Abstract: The introduction of transverse velocity fluctuations into a separated shear layer on an airfoil at high angles of attack is presently demonstrated to be an effective separation-control technique. Airfoil aerodynamic characteristics, including poststall lift and drag as well as maximum lift coefficient and stall angle, all exhibited improvements controlled forcing at 20 deg angle of attack led to an increased spreading of the mean velocity profile, together with increased turbulence activity; separation moved from the leading edge to about 80 percent of chord.

Vorticity transport and wake structure for bluff bodies at finite Reynolds number

Abstract: This paper is concerned with the structure of recirculating wakes downstream of bluff bodies at finite Reynolds number, and the relationship between these wakes and asymptotic behavior for R→∞. It is shown that recirculating eddies at finite Reynolds number are not necessarily a precursor of the asymptotic phenomenon of separation that is ‘‘predicted’’ via boundary‐layer theory. In some cases, such as the flow around deformable gas bubbles, recirculating eddies may appear at intermediate Reynolds numbers, but then disappear with further increase of the Reynolds number. In other cases, recirculating eddies exist downstream of the body at finite Reynolds number that are completely detached from the body. These examples suggest that recirculating eddies at large, but finite Reynolds number do not necessarily imply anything about the solution behavior for R→∞, even for solid bodies where it is usually assumed that wakes for R∼O(10–100) are a manifestation of the same processes that lead to separation of bound...

Frontiers in experimental fluid mechanics

Abstract: 1. The Influence of Developments in Dynamical Systems Theory on Experimental Fluid Mechanics.- 2. Low Speed, lndraft Wind Tunnels.- 3. High-Reynolds Number Liquid Flow Measurements.- 4. The Turbulent Boundary Layer.- 5. The Art and Science of Flow Control.- 6. Microbubble Drag Reduction.- 7. Unsteady Pulsing of Cylinder Wakes.- 8. Vortex Dynamics of Delta Wings.- 9. Accomplished Insect Fliers.- 10. The Aeroacoustics of Trailing Edges.

Crossflow-vortex instability and transition on a 45 deg swept wing

Abstract: A crossflow vortex experiment on a 45 deg swept wing is currently being conducted in the Arizona State University Unsteady Wind Tunnel. The experimental apparatus is designed to produce crossflow-dominated transition by simulating infinite swept wing flow using contoured end liners in a closed throat wind tunnel. Stationary fixed-wavelength crossflow vortices are observed at several chord Reynolds numbers. The vortex wavelength which is fixed for a given Reynolds number varies with Reynolds number approximately as predicted by linear stability theory, but with the predicted wavelengths about 30 percent larger than the observed wavelengths. Travelling waves are observed both in the frequency range predicted by linear stability theory and at higher frequencies. These higher frequency waves may be harmonics of the primary crossflow waves generated by a nonlinear parametric resonance phenomena.

On the breakdown of the wave packet trailing a turbulent spot in a laminar boundary layer

Abstract: The evolution of two oblique wave packets trailing, a transitional spot in a laminar boundary layer was investigated in order to determine the extent of the interaction between the packets and the spot. The experimental investigation, carried out on two slightly different laminar boundary layers characterized by Falkner-Skan constants of β = 0 and β = 0.2, revealed that very small pressure gradients can have significant effects on the stability of the laminar boundary layer and the rate at which it is contaminated by a turbulent spot. Some simple, novel statistical procedures for treating the data were developed and were used to accentuate the understanding of the physical processes governing transition to turbulence.

Boundary-layer predictions for small low-speed contractions

Abstract: The present scheme for the prediction of boundary-layer development in small, low-speed wind tunnel contraction sections proceeds by calculating the wall pressure distributions, and hence the wall velocity distributions, by means of a three-dimensional potential-flow method. For the family of contractions presently treated, the assumption of a laminar boundary layer appears to be justified; the measured boundary layer momentum thicknesses at the exit of the four contractions were found to lie within 10 percent of predicted values.

Prediction of Transition Due to Isolated Roughness

Abstract: A combination of an interactive boundary-layer procedure and ^"-method has been used to determine the locations of the onset of transition for flows over a flat plate with bumps or hollows. The details of the procedure are examined and the need for fine-grid calculations in regions of strong adverse pressure gradients emphasized. The location of initial disturbance is shown to be unimportant, provided it does not occur in the region where the critical Reynolds number is higher than the flow Reynolds number. The procedure allows good results for flows with and without separation. Possible means of improvement of the method are discussed.

Effect of stepheight on the separated flow past a backward facing step

Abstract: This Brief Communication presents a numerical study of planar laminar flow over a backward facing step. The principal parameters are the expansion ratio (defined as the ratio of stepheight to channel height) and the Reynolds number (based on the maximum inflow velocity and the hydraulic diameter of the channel inlet). Results are presented for expansion ratios of 0.25–0.75 and Reynolds numbers from 50–900. The computed results are in close agreement with previous experimental and numerical results for expansion ratios of 0.5. The effect of the expansion ratio on the reattachment length is summarized as a function of the Reynolds number.

The relaxation of a turbulent boundary layer in an adverse pressure gradient

Abstract: Reattached turbulent boundary layer relaxation downstream of a wall fence is investigated. An adverse pressure gradient is imposed upon it which is adjusted to bring the boundary layer into equilibrium. The pressure gradient is adjusted so as to bring the Clauser parameter G down to a value of about 11.4 and then maintain it constant. In the region from the reattachment point to 2 or 3 reattachment lengths downstream, the boundary layer recovers from the initial major effects of reattachment. Farther downstream, where G is constant, the pressure-gradient parameter changes very slowly and profiles of non-dimensionalized eddy viscosity appear self-similar. However, pressure gradient and eddy viscosity are both roughly twice as large as expected on the basis of previous equilibrium turbulent boundary layer studies.

The delay of turbulent boundary layer separation by oscillatory active control

Abstract: : Flow over a solid wedge from which a fully developed, turbulent boundary layer separates naturally, was investigated experimentally. The flow, which separates at the geometric discontinuity, turns into a free mixing layer downstream of it. However the flow can be forced to reattach by introduction of two-dimensional, harmonic, and small-amplitude perturbations at the apex of the wedge. The temporally-averaged characteristics of the reattached boundary layer are typical to flows at less severe pressure gradients, with the exception of the spanwise coherence near the solid surface which has been notably enhanced by the imposed perturbations. Phase-locked and ensemble-averaged results indicate that the subharmonic frequency dominates the flow at large distances from the apex at all forcing frequencies considered thus far. The preliminary results which are presented indicate that this might be an effective way to delay separation of turbulent as well as laminar boundary layers. Keywords: Boundary layer flow; Flow separation; Boundary layer control; Stalling. Israel.

Turbulent flows with incipient separation over solid waves

Abstract: Measurements of the streamwise velocity over solid sinusoidal waves with height to wavelength ratios of 2a/λ=0.05, 0.125 and with dimensionless wave numbers α+=0.00624, 0.00135 have been made. For these conditions the instantaneous flow reverses direction, but the time-averaged flow is non-separated. Many features of the flow are similar to those reported in previous papers for a time-averaged flow that is separated. Approximate agreement is obtained from an eddy viscosity model derived for flow over small amplitude waves. However, the differences are more interesting than the agreement in that they point out shortcomings of present Reynolds stress models. Comparison with other measurements in the literature shows how increasing Reynolds number decreases the size of the separated region. Measurements of pressure profiles and of drag are interpreted in terms of measured flow patterns.

Vortex-induced boundary-layer separation

Abstract: : Unsteady boundary-layer separation at high Reynolds numbers, Re, is considered on a theoretical and computational basis Whenever an external inviscid flow induces a region of adverse pressure gradient near a wall, the development of recirculating eddies in the boundary layer is common An unsteady viscous-inviscid interaction often follows in the form of a local boundary-layer eruption and abrupt ejection of near-wall vorticity into the external flow The dynamics of this process, as it develops in an initially thin boundary layer, is considered As interaction ensues, the flow focuses into a band which progressively narrows in the streamwise direction The complex flow development is extremely difficult to resolve using conventional Eulerian methods; here the boundary-layer solutions are obtained using Lagrangian methods, wherein trajectories of a large number of fluid particles are computed The algorithms developed are general but are applied here to the problem of the boundary-layer induced by a two-dimensional vortex above an infinite plane wall Solutions are obtained for the limit problem Re approaches limit of infinity, and for Re large but finite using an interacting boundary-layer approach The present results describe the initial stages of a strong unsteady viscous-inviscid interaction; apparently it is necessary to account for the effect of normal pressure variations to continue the interaction Keywords: Vortices; Turbulent boundary layer; Turbulent bursts; Boundary-layer eruptions; Unsteady interactions; Boundary layer flow separation

The delay of turbulent boundary layer separation by oscillatory active control

Abstract: The flow outside a solid wedge that abruptly diverges at an angle of 18° was investigated experimentally. A turbulent boundary layer, which separated at the discontinuity and turned into a mixing layer downstream of it, reattached as a result of harmonic excitation at the apex of the wedge. Preliminary results indicate that this might be an effective way to delay separation of turbulent and laminar boundary layers.

Calculation of convective heat transfer in recirculating turbulent flow using various near-wall turbulence models

Abstract: The turbulent flow about a two-dimensional blunt rectangular section is used as a test case to examine the performance of seven near-wall turbulence models The first two models are one-equation low Reynolds number (Re) models requiring a fine grid near-wall treatment The other near-wall turbulence models considered are based on wall functions that bridge with a single cell the thermally important near-wall region Standard wall functions based on the local equilibrium assumption, wall functions using a two- and three-layer approach to evaluate local variations of turbulence quantities in the k equation, and extension of the two- and three-layer treatments to the ∊ equation are considered The numerical predictions are obtained using a variant of the k-∊ model incorporating a curvature correction The governing equations are discretized using a finite-volume formulation employing the bounded-skew hybrid differencing scheme The solutions are obtained using a two-pass procedure, devised to allow for the c

Separation shock dynamics in Mach 5 turbulent interactions induced by cylinders

Abstract: Wall pressure fluctuations have been measured under the unsteady separation shock in interactions generated by unswept circular cylinders. Models were tested in the turbulent boundary layers on the tunnel floor and on a full-span flat plate. The freestream unit Reynolds number was 53×10 6 m -1 , and the wall temperature approximately adiabatic. Distributions of the shock frequency, shock period, and shock speeds in the upstream and downstream directions have been calculated

Analysis of low Reynolds number separation bubbles using semiempirical methods

Abstract: The formation and growth of transitional separation bubbles can significantly affect boundary-layer development on airfoils operating at low chord Reynolds numbers. Of primary concern is the change in boundary-layer thickness between laminar separation and turbulent reattachment. This can be estimated using semiempirical methods, such as the one devised by Horton (1968), which are based on solutions to the integral forms of the boundary-layer equations. The applicability of these methods at low Reynolds numbers was investigated using hot-wire measurements of bubbles formed on an NACA 66(3)-018 airfoil at chord Reynolds numbers of 50,000-200,000. The momentum thickness growth between separation and transition was found to be similar to that predicted for a laminar half-jet and appears to be influenced by the momentum thickness Reynolds number at separation. This parameter also was found to have a noticeable effect on the Reynolds number based on the length of a bubble's laminar portion.