SEARCH · Engineering Papers
Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.
Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.
Experimental work with leading edge separation bubbles is presented to clarify the issues regarding transition in separated regions. Hot-wire measurements, in the form of oscilloscope traces, turbulence intermittency and conditionally sampled velocity distributions are given. The resulting points of transition onset and spot production rates are compared to existing correlations.
An inverse boundary layer procedure for calculating separated, turbulent boundary layers at infinitely long, crabbing wing was developed. The procedure was developed for calculating three dimensional, incompressible turbulent boundary layers was expanded to adiabatic, compressible flows. Example calculations with transsonic wings were made including viscose effects. In this case an approximated calculation method described for areas of separated, turbulent boundary layers, permitting calculation of this displacement thickness. The laminar boundary layer development was calculated with inclined ellipsoids.
A detailed Investigation of a flow in which a three-dimensional shock wave separates a two-dimensional turbulent boundary layer is presented. The resulting flowfield is highly three dimensional with a significant portion of flow separation on the surface at the phi = 0 deg (windward) plane was well as a large zone of secondary surface flow off this plane. Mean and fluctuating experimental measurements were obtained throughout the entire flowfield. These measurements included mean pressures, flow angles and shear on the surface, as well as yaw angles, static pressures, turbulent shear stresses, and turbulent kinetic energies on selected planes throughout the flowfield. In addition, numerical predictions of this flow, obtained by solving the Navier-Stokes equations with an algebraic eddy viscosity turbulence model, are presented. These computations reasonably predict both the surface and flowfield quantities, despite the extremely complicated nature of the experimental flow.
Planar Particle Image Velocimetry measurements were obtained in the separating/reattaching flow region downstream of an axisymmetric backward-facing step. Data were acquired for a two-dimensional (2D) separating boundary layer at five different Reynolds numbers based on step height (Re(sub h)), spanning 5900-33000, and for a three-dimensional (3D) separating boundary layer at Re(sub h) = 5980 and 8081. Reynolds number effects were investigated in the 2D cases using mean-velocity field, streamwise and wall-normal turbulent velocity, and Reynolds stress statistics. Results show that both the reattachment length (x(sub r)) and the secondary separation point are Reynolds number dependent. The reattachment length increased with rising Re(sub h) while the secondary recirculation region decreased in size. These and other Re(sub h) effects were interpreted in terms of changes in the separating boundary layer thickness and wall-shear stress. On the other hand, in the 3D case, it was found that the imposed cross-flow component was relatively weak in comparison to the streamwise component. As a result, the primary influences of three dimensionality only affected the near-separation region rather than the entire separation bubble.
Experimental research has been conducted on the effects of wall cooling, Mach number, and unit Reynolds number on the transition Reynolds number of cylindrical separated boundary layers on an ogive-cylinder model. Results were obtained from pressure and temperature measurements and shadowgraph observations. The maximum scope of measurements encompassed Mach numbers between 2.06 and 4.24, Reynolds numbers (based on length of separation) between 60,000 and 400,000, and ratios of wall temperature to adiabatic wall temperature between 0.35 and 1.0. Within the range of tile present tests, the transition Reynolds number was observed to decrease with increasing wall cooling, increase with increasing Mach number, and increase with increasing unit Reynolds number. The wall cooling effect was found to be four times as great when the attached boundary layer upstream of separation was cooled in conjunction with cooling of the separated boundary layer as when only the separated boundary layer was cooled. Wall cooling of both the attached and separated flow regions also caused, in some cases, reattachment in the otherwise separated region. Cavity resonance present in the separated region for some model configurations was accompanied by a large decrease in transition Reynolds number at the lower test Mach numbers.
The results of experimental trials to obtain measurements of the effects of frequency on the sinusoidal variations in the free stream velocity over a separating turbulent boundary layer are reported. An oscillation frequency of 0.596 Hz was examined, in conjunction with a free stream velocity of 10.18 m/sec. Hot-wire anemometers measured the velocity upstream of the separation and the phase-averaged skin friction, while the velocities in the detached flow zone and the downstream velocity were assayed with a laser anemometer. Large amplitude and phase variations developed at the beginning of separation, producing hysteresis in the relationships between flow parameters. Increases in the free stream velocity decreased the detached shear layer thickness, while in the deceleration segment of the cycle the separated shear layer grew thicker. The results are in a frequency range that are significant for axial compressor blades and helicopter blades.
Several multiequation eddy viscosity models of turbulence are used with the Navier-Stokes equations to compute three classes of experimentally documented shock-separated turbulent boundary-layer flows. The types of flow studied are: (1) a normal shock at transonic speeds in both a circular duct and a two-dimensional channel; (2) an incident oblique shock at supersonic speeds on a flat surface; and (3) a two-dimensional compression corner at supersonic speeds. Established zero-equation (algebraic), one-equation (kinetic energy), and two-equation (kinetic energy plus length scale) turbulence models are each utilized to describe the Reynolds shear stress for the three classes of flows. These models are assessed by comparing the calculated values of skin friction, wall pressure distribution, velocity, Mach number, and turbulent kinetic energy profiles with experimental measurements. Of the models tested the two-equation model results gave the best overall agreement with the data.
Measurements were obtained of the sinusoidal unsteadiness of the free stream velocity during the separation of the turbulent boundary layer. Data were gathered by single wire and cross-wire, anemometry upstream of flow detachment, by laser Doppler velocimetry to detect the movement of the flow in small increments, and by a laser anemometer in the detached zone to measure turbulence and velocities. The study was restricted to a sinusoidal instability frequency of 0.61 and a ratio of oscillation amplitude to mean velocity of 0.3. Large amplitude and phase variations were found after the detachment, with unsteady effects producing hysteresis in the relationships between flow parameters. The detached shear layer decreased in thickness with increasing free-stream velocity and increases in the Reynolds shear stress. Deceleration of the free stream velocity caused thickening in the shear layer and upstream movement of the flow reversal location. The results are useful for studies of compressor blade and helicopter rotors in transition.
Instability and transition to turbulence in separated shear layer produced by laminar boundary layer separation from rearward-facing step, noting periodic spanwise structure
Classical boundary layer theory is inadequate to deal with the problem of flow separation owing to its underlying assumption that the boundary layer has an insignificant effect on the external stream. This difficulty is resolved by a theory which includes interaction with the external flow. This theory is described from the viewpoint of the asymptotic triple deck structure. Several triple deck studies are reviewed with emphasis on results of interest in aeronautical applications.
Following the law of stress adopted in the Navier-Stokes equations, the configuration of the viscous flow in planes at right angles to the axis of an infinite cylinder is found to be independent of the axial motion of the cylinder. In the limiting case of a yawed or swept wing of very high aspect ratio, certain boundary-layer and separation phenomena are thus determined independently by the crosswise component of velocity.
Following a law of stress adopted in the Navier-Stokes equations, the configuration of the viscous flow in planes at right angles to the axis of an infinite cylinder is found to be independent of the axial motion of the cylinder. In the limiting case of a yawed or swept wing of very high aspect ratio, certain boundary-layer and separation phenomena are thus determined independently by the crosswise component of velocity. It follows that the effect of sweepback is to increase the area of stable laminar flow and to decrease the lift coefficient at which flow separation occurs.
A full 3D Navier-Stokes numerical investigation has been conducted of the shock-wave/boundary-layer flow interactions caused by four diamond-shaped struts, of varying thickness, in an annular duct with Mach 3 core flow and turbulent boundary-layers on both walls. Secondary flows caused by weak-to-strong interactions are examined in the vicinity of a strut which is bounded by curved endwalls. The duct endwall boundary-layer separated for the strongest interaction. The struts studied had maximum thickness-to-chord ratios of 0.125, 0.188, 0.250, and 0.500. The duct gap height is 0.7 strut chords, the duct inner-to-outer wall radius ratio is 0.7, and the Reynolds number is 3 x 10 exp 5 based on the strut chord length which was held constant for all interactions considered. The effects of strut thickness on the secondary flows are discussed, including: trajectories for the leading and trailing edge horseshoe vortices, strut/endwall corner vortices, and boundary-layer separation. The line of coalescence discussed in the literature, previously ascribed to boundary-layer separation, is shown to be caused by the leading edge horseshoe vortex convecting along the shock front.
Numerical solutions of the complete, time-averaged conservation equations using several eddy-viscosity models for the Reynolds shear stress to close the equations are compared with experimental measurements in a compressible, turbulent separated flow. An efficient time-splitting, explicit difference scheme was used to solve the two-dimensional conservation equations. The experiment used for comparison was a turbulent boundary layer that was separated by an incident shock wave in a Mach 2.93 flow with a unit Reynolds number of 5.7 x 10 to the seventh power m. Comparisons of predicted and experimental values of surface pressure, shear stress along the wall, and velocity profiles are shown. One of the tested eddy-viscosity models which allows the shear stress to be out of equilibrium with the mean flow produces substantially better agreement with the experimental measurements than the simpler models. A tool is thereby provided for inferring additional information about the flow, such as static pressures in the stream, which might not be directly obtainable from experiments.
A finite difference boundary layer algorithm was developed to model viscous effects when an inviscid core flow solution is given. This algorithm solved each boundary layer equation separately, then iterated to find a solution. Solving the boundary layer equations sequentially was 2.4 to 4.0 times faster than solving the boundary layer equations simultaneously. This algorithm used a modified Baldwin-Lomax turbulence model, a weighted average of forward and backward differencing of the pressure gradient, and a backward sweep of the pressure. With these modifications, the boundary layer algorithm was able to model flows with and without separation. The number of grid points used in the boundary layer algorithm affected the stability of the algorithm as well as the accuracy of the predictions of friction coefficients and momentum thicknesses. Results of this boundary layer algorithm compared well with experimental observations of friction coefficients and momentum thicknesses. In addition, when used interactively with an inviscid flow algorithm, this boundary layer algorithm corrected for viscous effects to give a good match with experimental observations for pressures in a supersonic inlet.
A finite difference boundary layer algorithm was developed to model viscous effects when an inviscid core flow solution is given. This algorithm solved each boundary layer equation separately, then iterated to find a solution. Solving the boundary layer equations sequentially was 2.4 to 4.0 times faster than solving the boundary layer equations simultaneously. This algorithm used a modified Baldwin-Lomax turbulence model, a weighted average of forward and backward differencing of the pressure gradient, and a backward sweep of the pressure. With these modifications, the boundary layer algorithm was able to model flows with and without separation. The number of grid points used in the boundary layer algorithm affected the stability of the algorithm affected the stability of the algorithm as well as the accuracy of the predictions of friction coefficients and momentum thicknesses. Results of this boundary layer algorithm compared well with experimental observations of friction coefficients and momentum thicknesses. In addition, when used interactively with an inviscid flow algorithm, this boundary layer algorithm corrected for viscous effects to give a good match with experimental observations for pressures in a supersonic inlet.
Building block experiments and companion numerical simulations intended to verify and guide turbulence modeling are described. A series of experiments and computations being used to enhance modeling development for the shock wave turbulent boundary layer interaction problem is emphasized. Results are given for transonic flow over a circular arc airfoil undergoing shock wave induced, boundary layer separation for supersonic flow along a tube wall undergoing normal shock wave induced, boundary layer separation. Experimental data which use the complete Navier-Stokes equations are discussed.
The properties of a Navier-Stokes solution of a shock-separated turbulent flow over a flat wall are investigated. Refinements of an algebraic relaxation turbulence model previously shown to be of value for the simulation of separated flows are presented. A simplified analysis applicable near an adiabatic wall is developed and used to help verify the accuracy of the numerical solution. Features of the time-dependent response of a turbulent boundary layer to shock impingement are presented.