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.
Fluid dynamic aspects of the Antarctic ozone hole phenomena are studied. Data collected by the ER-2 aircraft as part of the Airborne Antarctic Ozone Experiment (AAOE) are used to calculate the potential vorticity distribution on potential temperature surfaces. Most of the ER-2 flights show a monotonic decrease in potential vorticity and nitrous oxide toward the pole on isentropic surfaces.
The passage of a weak vorticity disturbance through a reactive shock wave, or detonation, is examined by means of a linearized treatment. Of special interest is the effect of chemical heat release on the amplification of vorticity in particular, and on the disturbance pattern generated downstream of the detonation in general. It is found that the effect of exothermicity is to amplify the refracted waves. The manner in which the imposed disturbance alters the structure of the detonation itself is also discussed.
The nonlinear interaction between planar or near-planar Tollmien-Schlichting waves and longitudinal vortices, induced or input, is considered theoretically for channel flows at high Reynolds numbers. Several kinds of nonlinear interaction, dependent on the input amplitudes and wavenumbers or on previously occurring interactions, are found and inter-related. The first, Type 1, is studied the most here and it usually produces spanwise focusing of both the wave and the vortex motion, within a finite scaled time, along with enhancement of both their amplitudes. This then points to the nonlinear interaction Type 2 where new interactive effects come into force to drive the wave and the vortex nonlinearly. Types 3, 4 correspond to still higher amplitudes, with 3 being related to 2, while 4 is connected with a larger-scale interaction 5 studied in an allied paper. Both 3, 4 are subsets of the full three-dimensional triple-deck-lie interaction, 6. The strongest nonlinear interactions are those of 4, 5, 6 since they alter the mean-flow profile substantially, i.e., by an 0(1) relative amount. All the types of nonlinear interaction however can result in the formation of focussed responses in the sense of spanwise concentrations and/or amplifications of vorticity and wave amplitude.
The receptivity problem for Goertler vortices induced by wall roughness is investigated. The roughness is modelled by small amplitude perturbations to the curved wall over which the flow takes place. The amplitude of these perturbations is taken to be sufficiently small for the induced Goertler vortices to be described by linear theory. The roughness is assumed to vary in the spanwise direction on the boundary layer lengthscale, while in the flow direction the corresponding variation is on the lengthscale over which the wall curvature varies. In fact the latter condition can be relaxed to allow for a faster streamwise roughness variation so long as the variation does not become as fast as that in the spanwise direction. The function which describes the roughness is assumed to be such that its spanwise and streamwise dependences can be separated; this enables progress by taking Fourier or Laplace transforms where appropriate. The cases of isolated and distributed roughness elements are investigated and the coupling coefficient which relates the amplitude of the forcing and the induced vortex amplitude is found asymptotically in the small wavelength limit. It is shown that this coefficient is exponentially small in the latter limit so that it is unlikely that this mode can be stimulated directly by wall roughness. The situation at 0(1) wavelengths is quite different and this is investigated numerically for different forcing functions. It is found that an isolated roughness element induces a vortex field which grows within a wedge at a finite distance downstream of the element. However, immediately downstream of the obstacle the disturbed flow produced by the element decays in amplitude. The receptivity problem at larger Goertler numbers appropriate to relatively large wall curvature is discussed in detail.
The problem of Goertler vortices in compressible boundary layers over concave walls is considered. At O(1) wavelengths, the instability is governed by parabolic partial differential equations that are solved numerically to determine the effect of various initial conditions on the development of Goertler vortex instability in compressible boundary layers. The results show that both the velocity and temperature fluctuations may lead to a Goertler vortex. The vortex growth rates determined from the present method are found to differ somewhat from those given by a normal mode solution. At both the supersonic and hypersonic Mach numbers, cooling has a small destabilizing effect. In addition, the most unstable disturbances shift toward lower wavelengths because of thinning of the boundary layer. The results also show that compressibility has a stabilizing effect on the Goertler instability, while the effect of an adverse pressure gradient is found to be destabilizing. The behavior of the Goertler vortex structure with Mach number is also examined. At hypersonic Mach numbers, vortices are located near the edge of the boundary layer for adiabatic wall conditions. However, the entire boundary layer is affected when the wall is cooled.
The latest stages of work on the interaction between longitudinal vortices and turbulent boundary layers show that very large changes in turbulence structure occur when the vortices are strong (crossflow angles of order 20 deg). The changes are poorly correlated by current turbulence models and go well beyond the rotation of the stress tensor in the vortex region that is explicitly represented by the exact 'generation' (exchange) terms in the Reynolds-stress transport equations. Measurements in the interaction between a burst vortex and a boundary layer show qualitatively similar results to the unburst case, but shed useful light on the bursting process itself.
This paper presents experimental data taken in the forward region of a separated internal free-shear layer produced in an internal cavity flowfield. It has been found that in the region very near the forward restrictor, experimental velocity profiles agree closely with the exact Stuart instability velocity profile with various values of a steepness parameter. Reynolds shear-stress profiles suggest the presence of counter-rotating longitudinal vortices. Spectral analysis by the maximum entropy method of the time samples within the vortices indicates subharmonic and harmonic components of the fundamental frequency, with a weak indication of the fundamental frequency itself.
Steady and unsteady, supersonic asymmetric vortical flows and their passive control around circular and noncircular cones are considered in this paper. These problems are formulated by using the unsteady, compressible, single and double, thin-layer. Navier-Stokes equations. The equations are solved by using an implicit, upwind, flux-difference splitting, finite-volume scheme, either in a pseudotime stepping or in an accurate-time stepping. An implicit, approximately-factored, central-difference finite-volume scheme has also been used to validate some applications of the upwind scheme. Steady asymmetric vortical flows have been predicted by using random and controlled disturbances for circular and noncircular cones. Unsteady asymmetric vortex-shedding flows have also been predicted, for the first time, using time-accurate solutions, for circular and noncircular cones. Control of flow asymmetry have been demonstrated computationally, for the first time, by inserting a vertical fin the leeward plane of geometric symmetry.
Navier-Stokes solutions of subsonic vortical flow over a 75 deg sweep delta wing with a sharp leading edge are presented. The sensitivity of the solution to the numerical scheme is examined using both a partially upwind scheme and a scheme with central differencing in all directions. At moderate angles of attack, no vortex breakdown is observed, whereas the higher angle-of-attack cases exhibit breakdown. The effect of numerical grid density is investigated, and solutions that are obtained with various grid densities are compared with experimental data. An embedded grid approach is implemented to enable higher resolution in selected isolated flow regions, such as the leeward-side surface, the leading-edge vortical flow, and the vortex breakdown region.
The velocity field induced by a vorticity distribution decaying rapidly in the distance from the origin is investigated. It is shown that the sum of vector potentials for the velocity field can be expressed as a linear combination of Mn values, where Mn denotes the number of linearly independent vector functions of nth order. It is then shown that only the linear combinations 2n + 1 of these Mn vector functions contribute to the far field velocity which is irrotational, and that the corresponding scalar potential is then represented by a linear combination of 2n + 1 spherical harmonics of nth order whose coefficients are linear combinations of nth moments of vorticity.
The passage of a weak vorticity disturbance through a reactive shock wave, or detonation, is examined by means of a linearized treatment. Of special interest is the effect of chemical heat release on the amplification of vorticity in particular, and on the disturbance pattern generated downstream of the detonation in general. It is found that the effect of exothermicity is to amplify the refracted waves. The manner in which the imposed disturbance alters the structure of the detonation itself is also discussed.
The theoretical generalization, synthesis, and classification of various relevant formulations of the vorticity-dilatation conditioning problem is presented. Kinematic and dynamic conditions as well as differential and integral forms are considered. The possibility of using various variables that relate directly to vorticity and dilatation to construct the desired boundary conditions is examined. Some new formulations are given and previously proposed ones are located properly in the classification, and their interrelations are clarified.
Using standard data analysis techniques, researchers explore the links between disturbance growth and quasi-geostrophic potential vorticity (PV) gradients; appearance and disappearance of cutoff lows and blocking highs and their relation to a zonal index (properly defined in terms of PV); and teleconnections between different flow patterns and their relation to the zonal index. It was found that the PV index and the eddy index correlate better than a zonal index (defined by zonal wind) and the eddy index. In the frequency domain there are three frequencies (.03, .07 and .17 cpd (cycle per day) corresponding to periods of 33, 14 and 6 days) at which PV index and the eddy index exhibit local maxima. The high correlation found at periods of 33 days is mainly due to eddy activity at high latitudes while the local correlation maxima found at the shorter periods are mainly due mid-latitude eddy activity. The correlation between the PV index and the geopotential height anomaly at 500 mb, at each grid point in the Northern Hemisphere, shows the existence of most of the teleconnection patterns summarized by Wallace and Gutzler (1981): the North Atlantic Oscillation, the North Pacific Oscillation, and the Pacific/North American patterns. Results show that the Isentropic Potential Vorticity (IPV) analysis can be a very useful and powerful tool when used to understand the dynamics of several large scale atmospheric systems. Although the data are limited to only one winter, and it is difficult to assess the statistical significance of the correlation coefficients presented here, the results are encouraging from physical viewpoint.
A summary of flow visualization data obtained as part of NASA Grant NAG2-258 is presented. During the course of this study, many still and high speed motion pictures were taken of the leading edge vortices on a series of flat plate delta wings at varying angles of attack. The purpose is to present a systematic collection of photographs showing the state of vortices as a function of the angle of attack for the four models tested.
A review of the solution techniques of Tezduyar et al. (1988 and 1989) for the vorticity-streamfunction formulation of two-dimensional incompressible flows is presented. While both the viscous and inviscid cases are considered, the derivation of the proper finite-element formulations for multiply connected domains is emphasized. In all formulations associated with the vorticity transport equation, the streamline upwind/Petrov-Galerkin method is used. The adaptive implicit-explicit and grouped element-by-element solution strategies are employed to maximize the computational efficiency. The solutions obtained in all test cases compare well with solutions from previously published investigations. The convergence and benchmark studies performed in this paper show that the solution techniques presented are accurate, reliable, and efficient.
A method for calculating the incompressible viscous flow about two-dimensional bodies, utilizing the velocity-vorticity form of the Navier-Stokes equations using a staggered-grid formulation is presented. The solution is obtained by employing an alternative-direction implicit method for the solution of the block tridiagonal matrix resulting from the finite-difference representation of the governing equations. The boundary vorticity and the conservation of mass are calculated implicitly as a part of the solution. The mass conservation is calculated to machine zero for the duration of the computation. Calculations for the flow about a circular cylinder, a 2-pct thick flat plate at 90-deg incidence, an elliptic cylinder at 45-deg incidence, and a NACA 0012, with and without a deflected flap, at - 90-deg incidence are performed and compared with the results of other numerical investigations.
Eastward and westward traveling waves were observed by the Nimbus 7 Limb Infrared Monitor of the Stratosphere (LIMS) during the northern winter 1978-1979. Eastward waves were prevalent in early winter and were involved in a minor Canadian warming in December 1978. A large westward traveling wave, as described by previous authors, was observed in January 1979 during a series of minor warmings. By comparing these two events, it is shown that in both cases the superposition of traveling and quasi-stationary waves led to constructive interference that was responsible for the warmings. However, there was significant asymmetry between eastward and westward traveling components. A local Eulerian analysis of potential vorticity (PV) transport indicates that adiabatic, geostrophic advection by the resolvable scales of motion explains qualitatively (but not quantitatively) the observed potential vorticity tendencies in the LIMS Northern Hemisphere winter. In particular, calculated advection explains the eastward rotation of the main vortex, intrusion of low PV air into the polar cap, and formation of high PV filaments at the vortex periphery.
The nonlinear interaction between planar or near-planar Tollmien-Schlichting waves and longitudinal vortices, induced or input, is considered theoretically for channel flows at high Reynolds numbers. Several kinds of nonlinear interaction, dependent on the input amplitudes and wavenumbers or on previously occurring interactions, are found and inter-related. The first, Type 1, is studied the most here and it usually produces spanwise focusing of both the wave and the vortex motion, within a finite scaled time, along with enhancement of both their amplitudes. This then points to the nonlinear interaction Type 2 where new interactive effects come into force to drive the wave and the vortex nonlinearly. Types 3, 4 correspond to still higher amplitudes, with 3 being related to 2, while 4 is connected with a larger-scale interaction 5 studied in an allied paper. Both 3, 4 are subsets of the full three-dimensional triple-deck-lie interaction, 6. The strongest nonlinear interactions are those of 4, 5, 6 since they alter the mean-flow profile substantially, i.e., by an O(1) relative amount. All the types of nonlinear interaction, however, can result in the formation of focused responses in the sense of spanwise concentrations and/or amplifications of vorticity and wave amplitude.