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.
The application of Rose's (1970) analytical method to the study of a shock wave interacting with a turbulent layer on a blunted compression surface is described. Only those details of the method that specifically apply to the case under consideration are discussed.
An experimental and analytical study of the formation of oxygen atoms during the oxidation of methane and methane-hydrogen mixtures behind incident shock waves was carried out over the temperature range 1790-2584 K at reaction pressures between 1.2 and 1.7 atm. Oxygen atom levels were determined indirectly by measurement of emission from reaction of O with CO. On the basis of these data and ignition-delay data reported in the literature, a kinetic scheme for methane oxidation was assembled. The proposed kinetic mechanism, in general, predicts higher peak oxygen atom levels than the current oxidation mechanisms proposed by Bowman and Seery and by Skinner and his co-workers.
An experiment is described that tests and guides computations of the interaction of a shock wave with a turbulent boundary layer. Numerical solutions of the time-averaged Navier-Stokes equations for the entire flow field employing algebraic eddy viscosity and turbulent Prandtl number models for shear stress and heat flux are presented and used to illustrate the dependence of the computations on the particulars of the turbulence models. To guide modifications in the models, the mean flow profiles and surface measurements of pressure, shear, and heat flux are analyzed critically. The results show that the models of eddy viscosity require substantial modifications in the interaction region. Improved solutions employing the experimentally modified models are presented.
An experimental investigation was conducted to determine the flowfield inside a bleed slot used to control an oblique shock-wave and turbulent boundary-layer interaction. The slot was oriented normal to the primary flow direction and had a width of 1.0 cm (primary flow direction), a length of 2.54 cm, and spanned 16.5 cm. The approach boundary layer upstream of the interaction was nominally 3.0 cm thick. Two operating conditions were studied: M = 1.98 with a shock generator deflection angle of 6 deg and M= 2.46 with a shock generator deflection angle of 8 deg. Measurements include surface and flowfield static pressure, Pitot pressure, and total mass-flow through the slot. The results show that despite an initially two-dimensional interaction for the zero bleed-flow case, the slot does not remove mass uniformly in the spanwise direction. Inside the slot, the flow is characterized by two separation regions which significantly reduce the effective flow area. The upper separation region acts as an aerodynamic throat resulting in supersonic flow through much of the slot.
Development of a boundary layer correction for an analysis program previously used by Bauer et al. (1972) to study two-dimensional flow past transonic airfoils at off-design conditions where shocks which interact with the boundary layer appear. Using a method which combines conformal mapping with a finite-difference scheme due to Murman and Cole (1971), an analysis is made of the interaction between shock waves on a supercritical wing section and a turbulent boundary layer, assuming that the shocks are weak and that separation is insignificant. It is shown that relatively simple shock wave/boundary layer interactions of the kind envisioned in this study can be treated by determining the displacement thickness iteratively together with the flow and altering the profile accordingly. It is concluded that the proposed boundary layer correction appears to furnish an essential improvement in cases where both pronounced aft loading and a large enclosed supersonic region of flow are present.
The faint optical filaments in Tycho's supernova remnant appear to be emission from a shock front moving at 5600 km/s. The intensity of the hydrogen lines, the absence of forbidden lines of heavy elements in the spectrum, and the width of the filaments are explained by a model in which a collisionless shock wave is moving into partially neutral gas. The presence of the neutral gas can be used to set an upper limit of approximately 5 x 10 to the 47th power ergs to the energy in ionizing radiation emitted by a Type I supernova. The patchy neutral gas is probably part of the warm neutral component of the interstellar medium. The existing information on the remnant of SN 1006 indicates that its emission is similar in nature to that from Tycho's remnant.
The ultraviolet spectrum of a Balmer-line filament located just outside the main body of optical filaments in the Cygnus Loop confirms the nonradiative shock wave theory for its origin. This theory is extended to include the hydrogen two-photon continuum, and the possibility of slow electron-ion equilibration in the postshock gas is considered. The shock velocity is inferred from the H-alpha profile. Comparison of model calculations with optical and ultraviolet spectra favors Coulomb equilibration behind a 170 km/s shock over models with rapid equilibration due to plasma turbulence. Elemental abundances in the preshock gas are found to be typical of diffuse interstellar clouds. The ram pressure behind the shock is higher than the pressures inferred for bright optical filaments. It is suggested that the bright optical filaments are regions of thermally unstable cooling behind shocks faster than the shock velocities inferred from their optical and ultraviolet spectra.
Results are presented of a dimensional study and an experimental investigation of the pressure rise across a shock wave which causes separation of the boundary layer on a flat plate. The experimental part of the investigation was conducted at a Mach number of 3.03 for a Reynolds number range of 2 x 10 (sup) 6 to 19 x 10 (sup) 6. The available experimental data are compared with the predictions of the present study, and the significance of the results obtained is discussed relative to certain practical design problems.
Description of the salient characteristics of temporal variations of greater than or about equal to 100 keV/nucleon protons and alpha particles during a ten-hour period associated with the passage of an interplanetary shock wave. In addition to extending measurements to energies considerably below previous observations for energetic storm particle and shock-spike events, the discovery of a new low energy feature, referred to as 'postshock spikes', is reported. The typical width of these spikes is several minutes, but they occur several hours after the passage of the shock front. Evidence is presented that spikes depends on the energy per charge of the particles.
Asymptotic solutions are derived for the pressure distribution in the interaction of a weak normal shock wave with a turbulent boundary layer. The undisturbed boundary layer is characterized by the law of the wall and the law of the wake for compressible flow. In the limiting case considered, for 'high' transonic speeds, the sonic line is very close to the wall. Comparisons with experiment are shown, with corrections included for the effect of longitudinal wall curvature and for the boundary-layer displacement effect in a circular pipe.
The finite difference computation method was investigated for solving problems of interaction between a shock wave and a laminar boundary layer, through solution of the complete Navier-Stokes equations. This method provided excellent solutions, was simple to perform and needed a relatively short solution time. A large number of runs for various flow conditions could be carried out from which the interaction characteristics and principal factors that influence interaction could be studied.
Solutions of the Reynolds-averaged Navier-Stokes equations are presented and compared with experimental surface data for a series of hypersonic shock-wave/turbulent-boundary-layer interaction flows. The turbulence models used include algebraic and two-equation eddy-viscosity models developed for transonic and supersonic flows. Also several additional modifications to the two-equation model to account for compressibility effects are developed and used. Although the modifications improve the agreement with the experimental data, no single model or modification correctly predicts all the test cases.
The rms intensities of fluctuating mass flux and total temperature and their correlation coefficients are given for the case of an adiabatic, Mach 4, axisymmetric shock-wave boundary-layer interaction. Data were obtained upstream, within, and downstream of the interaction by the use of constant temperature hot-wire anemometer. Turbulence spectra and quantitative behavior from oscilloscope traces are shown at selected locations. The measurements indicate that certain frequencies of the turbulence are increased as a result of the interaction and that the mass flux and total temperature fluctuations remain highly correlated over most of the boundary layer throughout the interaction. The present data are also transformed to rms intensities of fluctuating static temperature and velocity and compared with existing data obtained in adiabatic flows.
Fluctuating surface pressure measurements have been made to investigate the effectiveness of boundary layer separators (BLS's) in reducing the fluctuating pressure loads produced by separated shock wave turbulent boundary layer interactions. Measurements have been made under unswept and swept compression corner interactions in a Mach 5 flow. BLS's fix the separation location and eliminate the large-amplitude, low-frequency fluctuating pressure loads upstream of the compression corners. The loads on the unswept compression corner face are reduced by as much as 59%. The BLS's also shift the mean pressure distribution on the unswept corner face in the streamwise direction. Results show that the loads on the corner face vary with the BLS height and the distance between the BLS and the compression corner. Suggestions for the optimum placement and the use of the BLS's are also made.
The report assesses the capability for numerical simulation of 2-D and 3-D shock wave laminar and turbulent boundary layer interactions Three fundamental configurations are considered: single fin, double fin, and hollow cylinder flare. Thirteen separate cases were examined by a distinguished international group of researchers using the Reynolds-averaged Navier-Stokes (RANS) equations with a wide range of turbulence models from zero equation to full Reynolds Stress Equation formulations. The report presents an extensive comparison of computations and experimental data, summarizes the results, and makes recommendations for future research.
The Reynolds-averaged Navier-Stokes equations are solved numerically for supersonic flow over a blunt fin mounted on a flat plate. The fin shock causes the boundary layer to separate. which results in a complicated, three-dimensional shock-wave and boundary-layer interaction. The computed results are in good agreement with the mean static pressure measured on the fin and the flat plate. The main features, such as peak pressure on the fin leading edge and a double peak on the plate. are predicted well. The role of the horseshoe vortex is discussed. This vortex leads to the development of high-speed flow and, hence, low-pressure regions on the fin and the plate. Different thicknesses of the incoming boundary layer have been studied. Varying the thicknesses by an order of magnitude shows that the size of the horseshoe vortex and, therefore, the spatial extent of the interaction are dominated by inviscid flow and only weakly dependent on the Reynolds number. Colored graphics are used to show details of the interaction flow field.
An experimental study is conducted of the fluid-thermal-structural interaction of a clamped compliant panel exposed to a three dimensional shock-wave/boundary-layer interaction (SWBLI) induced by a Mach-6 compression ramp with a spanwise nonuniform incoming boundary layer. The nonuniform boundary layer was produced by placing trips on one side of the upstream flat plate, resulting in largely turbulent flow on the tripped side and transitional flow on the untripped side. Measurements of the flowfield confirmed that the tripped boundary layer contained elevated levels of unsteadiness, and the SWBLI was observed to vary from attached to fully separated as the ramp angle was increased from 10◦ to 38◦; the separation region on the tripped side of the panel was noticeably smaller, showing the elevated turbulence levels of the tripped-side flow to remain relatively localized rather than diffusing across the whole model. Full-field, time-resolved panel deformations were measured using high-speed photogrammetry and the vibrational response at each compression angle was characterized. Although the measured modes conformed largely to those from classical clamped-plate theory, some skewing of the mode shapes was observed. IR thermography highlighted regions of the compliant region where elevated temperatures were likely to promote thermal softening effects to the transient panel response. The quasi-static deformation and stress field was used to characterize the internal stress factor of each mode and showed a meaningful relationship between transient panel response and stress contained within each mode: modes with antinodes lying in high-stress areas of the plate tended to exhibit increases in vibrational frequency and decreases in vibrational power, whereas the opposite was true for modes with antinodes in low-stress areas.
A theory of fringe localization in rapid-double-exposure, diffuse-illumination holographic interferometry was developed. The theory was then applied to compare holographic measurements with laser anemometer measurements of shock locations in a transonic axial-flow compressor rotor. The computed fringe localization error was found to agree well with the measured localization error. It is shown how the view orientation and the curvature and positional variation of the strength of a shock wave are used to determine the localization error and to minimize it. In particular, it is suggested that the view direction not deviate from tangency at the shock surface by more than 30 degrees.