Search NASA⌕ Search

SEARCH · Search NASA

Results for “normalizing flows”

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.

At least 379 records · Page 21
All sourcesNASA NTRSDOE OSTI
NASA NTRS2002

On the Motion of an Annular Film in Microgravity Gas-Liquid Flow

Three flow regimes have been identified for gas-liquid flow in a microgravity environment: Bubble, Slug, and Annular. For the slug and annular flow regimes, the behavior observed in vertical upflow in normal gravity is similar to microgravity flow with a thin, symmetrical annular film wetting the tube wall. However, the motion and behavior of this film is significantly different between the normal and low gravity cases. Specifically, the liquid film will slow and come to a stop during low frequency wave motion or slugging. In normal gravity vertical upflow, the film has been observed to slow, stop, and actually reverse direction until it meets the next slug or wave.

McQuillen, John B.↗
NASA NTRS2006

Thermal Dispersion Within a Porous Medium Near a Solid Wall

The regenerator is a key component to Stirling cycle machine efficiency. Typical regenerators are of sintered fine wires or layers of fine-wire screens. Such porous materials are contained within solid-waH casings. Thermal energy exchange between the regenerator and the casing is important to cycle performance for the matrix and casing would not have the same axial temperature profile in an actual machine. Exchange from one to the other may allow shunting of thermal energy, reducing cycle efficiency. In this paper, temperature profiles within the near-wall region of the matrix are measured and thermal energy transport, termed thermal dispersion, is inferred. The data show how the wall affects thermal transport. Transport normal to the mean flow direction is by conduction within the solid and fluid and by advective transport within the matrix. In the near-wall region, both may be interrupted from their normal in-core pattern. Solid conduction paths are broken and scales of advective transport are damped. An equation is presented which describes this change for a wire screen mesh. The near-wall layer typically acts as an insulating layer. This should be considered in design or analysis. Effective thermal conductivity within the core is uniform. In-core transverse thermal effective conductivity values are compared to direct and indirect measurements reported elsewhere and to 3D numerical simulation results, computed previously and reported elsewhere. The 3-D CFD model is composed of six cylinders in cross flow, staggered in arrangement to match the dimensions and porosity of the matrix used in the experiments. The commercial code FLUENT is used to obtain the flow and thermal fields. The thermal dispersion and effective thermal conductivities for the matrix are computed from the results.

Simon, T.↗
NASA NTRS1984

Externally driven magnetic reconnection

A model is presented in which externaly driven reconnection is simulated by solving the MHD equations in an initially plane current sheet. Magnetic reconnection in the earth's magnetotail is widely believed to be the direct cause of magnetospheric substorms. Both 2-D and 3-D versions of the model have been developed. It is postulated that connection in the tail is triggered by a local compression of the plasma sheet which results from an invasion of the solar wind into the magnetotail. Thus, the simulation is started by introducing flow from the lobes normal to the plasma sheet. When resistivity is generated in a local region of the neutral sheet, reconnection develops and magnetic energy is converted into plasma bulk flow. Although the driven reconnection model is highly simplified, it can aid in understanding many features of substorms in the tail; in particular, results show that rapid flows both earthward and tailward of the neutral line and the nightside substorm current system are natural consequences of driven magnetic reconnection.

Walker, R. J.↗
NASA NTRS1993

The ground vortex flow field associated with a jet in a cross flow impinging on a ground plane for uniform and annular turbulent axisymmetric jets

An experimental study was conducted of the impingement of a single circular jet on a ground plane in a cross flow. This geometry is a simplified model of the interaction of propulsive jet exhaust from a V/STOL aircraft with the ground in forward flight. Jets were oriented normal to the cross flow and ground plane. Jet size, cross flow-to-jet velocity ratio, ground plane-to-jet board spacing, and jet exit turbulence level and mean velocity profile shape were all varied to determine their effects on the size of the ground vortex interaction region which forms on the ground plane, using smoke injection into the jet. Three component laser Doppler velocimeter measurements were made with a commercial three color system for the case of a uniform jet with exit spacing equal to 5.5 diameters and cross flow-to-jet velocity ratio equal to 0.11. The flow visualization data compared well for equivalent runs of the same nondimensional jet exit spacing and the same velocity ratio for different diameter nozzles, except at very low velocity ratios and for the larger nozzle, where tunnel blockage became significant. Variation of observed ground vortex size with cross flow-to-jet velocity ratio was consistent with previous studies. Observed effects of jet size and ground plane-to-jet board spacing were relatively small. Jet exit turbulence level effects were also small. However, an annular jet with a low velocity central core was found to have a significantly smaller ground vortex than an equivalent uniform jet at the same values of cross flow-to-jet velocity ratio and jet exit-to-ground plane spacing. This may suggest a means of altering ground vortex behavior somewhat, and points out the importance of proper simulation of jet exit velocity conditions. LV data indicated unsteady turbulence levels in the ground vortex in excess of 70 percent.

Cavage, William M.↗
NASA NTRS1990

Current collection by a long conducting cylinder in a flowing magnetized plasma

The numerical techniques, the definitions, and the normalizations used in the simulations of plasma flow past a long conducting cylinder with a magnetic field along the cylinder axis are described. The effect of cross-field plasma transport on the current collection without any contribution to the current from the field-aligned motion of the plasma particles is highlighted. The electric fields in the structure create a two-cell circulation of the electrons near the cylinder. The cell in the wake region has negative potentials. A fan-shaped circulation cell forms around the cylinder and in this cell the potential is generally positive. The geometry and the size of this positive cell affect the current collection. The potential structure around the cylinder is examined, along with its effect on the current collection and its oscillatory behavior. The variation of the time-average current as a function of the relative motion between the plasma and the cylinder is also investigated.

Singh, Nagendra↗
NASA NTRS1995

On the Calculation of Anisotropic Extinction Coefficients for Rigid Fibrous Ceramic Insulations

The specific anisotropic extinction coefficient e* couples the effective Radiative properties of a fibrous insulation into the radiation diffusion equation. This coefficient can be calculated using Mie scattering theory if fiber diameters, refractive indices and fiber orientation distributions are known. In general, fiber orientation distributions are not readily accessible and past calculations have considered fibers as either randomly distributed or normal to the heat flow direction. In certain rigid fibrous ceramic insulations neither of these cases apply well and a simple procedure is described for approximating e* from values calculated for the random and normal orientation cases. The intrinsic error associated with this scaling procedure is investigated. Numerical computations for several test structures and fiber materials show the average error to be less than 5% for net heat flux and radiation conductivity calculations.

Marshall, Jochen↗
NASA NTRSDate not supplied

Characterization of Jugular Venous Blood Flow During Acute Fluid Shifts

INTRODUCTION Exposure to weightlessness induces a headward fluid shift and redistribution of fluids, resulting in increased internal jugular vein (IJV) cross-sectional area and altered IJV blood flow dynamics including stasis and retrograde flow. These findings may contribute to various risks of spaceflight, including thrombosis or other risks affected by cerebral venous blood flow. To date, our understanding of the cerebral venous outflow dynamics is limited to the left IJV assessed ~45and ~150days into spaceflight. Therefore, it is unknown how quickly the alterations in venous blood flow dynamics change as a result of weightlessness. The purpose of this study is to determine 1) if venous stasis is an immediate effect of weightlessness, and therefore a risk for short-duration missions, and 2) the effect of weightlessness on brachiocephalic venous flow. METHODS In this study, we will determine the effects of acute weightlessness during parabolic flight on right and left IJV cross-sectional area, as well as blood flow dynamics through the IJV and brachiocephalic veins in 12 healthy subjects. Baseline ultrasound images will be obtained on the ground prior to the parabolic flight in the seated and supine postures to determine normal changes in blood flow with posture-induced fluid shifts in a 1genvironment. Subjects will then undergo parabolic flight withrepeated,~20 second exposure to 0g. Each data collection session will include 2D ultrasonography to quantify IJV cross-sectional area and Doppler ultrasonography to characterize venous blood flow patterns in the IJV and brachiocephalic veins. RESULTS Data collection for this study is planned to take place in October, 2021.CONCLUSIONS This study will characterize the immediate venous response in the left and right IJV and brachiocephalic veins upon entry into weightlessness. The results from this parabolic flight study will be compared with previous data from International Space Station crewmembers collectedafter~45 and ~150 days into spaceflight to provide a more complete understanding of the temporal profile of changes in venous blood flow dynamics during weightlessness and whether these findings develop in veins other than the left IJV. Supported by the NASA Human Research Program.

K Marshall-Bowman↗
NASA NTRS2022

Linear Disturbance Amplification Over Blunted Flat Plates in High-Speed Flows

Modal and nonmodal instability characteristics of cylindrically blunted flat plates with varying leading edge radii are described for Mach 4 and Mach 6 freestream conditions. The selection of leading edge radii and freestream parameters is informed by experimental conditions. The investigation of this 2D problem provides a slow entropy layer swallowing which allows for an isolated development of perturbations seeded upstream within different wall-normal regions of the flow. At both Mach numbers, a decrease in modal instability amplification was seen as the leading edge radius was increased. Nonmodal analysis reveals amplifying perturbations in the boundary layer as well as in the entropy layer. The medium bluntness regime exhibits the strongest amplification of nonmodal disturbances that is nonmonotonic in character. Small amplitude boundary forcing at the plate surface or volumetric forcing at various wall-normal heights was used to account for receptivity effects. While wall forcing effectively induced modal instabilities, only an actuation above the boundary layer captured disturbances that amplify within the entropy layer. The optimal nonmodal theory’s entropy-layer disturbance evolution exhibited outstanding agreement with controlled forcing, including receptivity effects. The evolution of entropy-layer disturbances from the optimal nonmodal theory showed excellent agreement with the results of receptivity to controlled forcing. Therefore, the nonmodal optimal growth analysis may provide a useful as well as efficient technique to identify the complete disturbance spectrum in blunt hypersonic configurations, where both modal and nonmodal disturbances can amplify in the boundary-layer and entropy-layer regions.

Boundary layer transition↗
NASA NTRS1998

Incompressible Turbulent Wing-Body Junction Flow

The overall objective of this study is to contribute to the optimized design of fan bypass systems in advanced turbofan engines. Increasing the engine bypass ratios have provided a major boost in engine performance improvement over the last fifty years. An engine with high bypass ratio (11-16:1) such as the Advanced Ducted Propulsion (ADP) is being developed and is expected to provide an additional 25% improvement in overall efficiency over the early turbofans. Such significant improvements in overall efficiency would reduce the cost per seat mile, which is a major government and Industry challenge for the 21th century. The research is part of the Advanced Subsonic Technology (AST) program that involves a NASA, U.S. Industry and FAA partnership with the goal of a safe and highly productive global air transportation system. The immediate objective of the study is to perform numerical simulation of duct-strut interactions to elucidate the loss mechanisms associated with this configuration that is typical of advanced turbofan engines such as ADP. However, at present experimental data for a duct-strut configuration are not available. Thus, as a first step a wing-body junction flow would be studied and is the specific objective of the present study. At the outset it is to be recognized that while duct-strut interaction flow is similar to that of wing-body junction flows, there are some differences owing to the presence of a wall at both ends of the strut. Likewise, some differences are due to the sheared inflow (as opposed to a uniform inflow) velocity profile. It is however expected that some features of a wing-body junction flow would persist. Next, some of the salient aspects of the complex flow near a wing-body junction, as revealed by various studies reported in the literature will be reviewed. One of the principle characteristics of the juncture flow, is the presence of the mean flow components in a plane perpendicular to the direction of the oncoming free-stream flow. The lateral curvature of the wing/strat causes the oncoming turbulent layer to skew about am axis (x-axis) parallel to the plane (xz-plane) of the mean shear. This is the principle mechanism for the generation of secondary flow. Such skew-induced secondary flows are slow to be attenuated by Reynolds stresses. Additional contribution to the generation of secondary flow comes from anisotropies in Reynolds stresses. Upstream of the strut, the mean-vorticity is directed span wise (along the y-direction). The presence of secondary flow in the vicinity of the strut causes the vorticity to stretch around the obstacle in a horse-shoe shape, with each leg having a vorticity of the opposite sense. The blockage effect of the strut imposes a severe adverse pressure gradient on the oncoming turbulent shear layer, causing boundary layer separation ahead of the leading edge, resulting in a vortex that rolls up and flows downstream into the juncture region. The separation vortices trailing in the wake of the wing can alter the lift or drag characteristics of the surfaces downstream of the wing-body juncture. Likewise, on submarines, the wake flow behind the appendage can degrade the performance of the propeller located downstream. The complex nature of this flow is caused by the presence of all six components of Reynolds stresses. Devenport and Simpson report that in the vicinity of the horse-shoe vortex there is intense recirculation with turbulent stresses being much larger than those normally observed in turbulent flows. These features contribute to making this flow a challenge to predict numerically. Some of the past studies provide useful insights into this flow that would guide our numerical efforts. In measurements reported by Shabaka and Bradshaw, the eddy viscosity tensor is seen to be non-isotropic and has negative components in certain regions. In an effort to evaluate the closure assumptions of various turbulence models, Devenport and Simpson used their own extensive measurements in juncture flows around the nose of a wing-body junction. Measured values of mean-velocity and/or turbulence kinetic energy was used to predict the magnitude of the shear stress vector. Algebraic stress models performed the best followed by Cebeci-Smith eddy viscosity model. The flow is reported to be dominated by a pressure field produced by the wing and the velocity field generated by the horseshoe vortex that is wrapped around the junction between the wing and wall. Kubendran et al. conclude from an experimental study that the shape of leading edge of the wing as characterized by its slenderness ratio is a major factor in determining the flow fields in the juncture region. The more thinner the leading edge of the juncture, the weaker the horseshoe vortex is. Also, with a slender leading edge, the secondary flow in the juncture would be due mainly to the cross-stream gradients of Reynolds stresses rather than due to a lateral skewing of the shear layer.

Krishnamurthy, R.↗
NASA NTRS2022

Linear Disturbance Amplification Over Blunted Flat Plates in High-Speed Flows

Modal and nonmodal instability characteristics of cylindrically blunted flat plates with varying leading edge radii are described for Mach 4 and Mach 6 freestream conditions. The selection of leading edge radii and freestream parameters is informed by experimental conditions. The investigation of this 2D problem provides a slow entropy layer swallowing which allows for an isolated development of perturbations seeded upstream within different wall-normal regions of the flow. At both Mach numbers, a decrease in modal instability amplification was seen as the leading edge radius was increased. Nonmodal analysis reveals amplifying perturbations in the boundary layer as well as in the entropy layer. The medium bluntness regime exhibits the strongest amplification of nonmodal disturbances that is nonmonotonic in character. Small amplitude boundary forcing at the plate surface or volumetric forcing at various wall-normal heights was used to account for receptivity effects. While wall forcing effectively induced modal instabilities, only an actuation above the boundary layer captured disturbances that amplify within the entropy layer. The optimal nonmodal theory’s entropy-layer disturbance evolution exhibited outstanding agreement with controlled forcing, including receptivity effects. The evolution of entropy-layer disturbances from the optimal nonmodal theory showed excellent agreement with the results of receptivity to controlled forcing. Therefore, the nonmodal optimal growth analysis may provide a useful as well as efficient technique to identify the complete disturbance spectrum in blunt hypersonic configurations, where both modal and nonmodal disturbances can amplify in the boundary-layer and entropy-layer regions.

Boundary layer transition↗
NASA NTRS1981

Semi-actuator disk theory for compressor choke flutter

A mathematical anaysis predict the unsteady aerodynamic utilizing semi actuator theory environment for a cascade of airfoils harmonically oscillating in choked flow was developed. A normal shock is located in the blade passage, its position depending on the time dependent geometry, and pressure perturbations of the system. In addition to shock dynamics, the model includes the effect of compressibility, interblade phase lag, and an unsteady flow field upstream and downstream of the cascade. Calculated unsteady aerodynamics were compared with isolated airfoil wind tunnel data, and choke flutter onset boundaries were compared with data from testing of an F100 high pressure compressor stage.

Micklow, J.↗
NASA NTRS1981

Measurements in ducted flows by laser-Doppler anemometry

The merits of small-scale, water-flow experiments with measurements by laser-Doppler anemometry are indicated in relation to previously reported results. New measurements are reported in a square-to-round transition duct and show that, for example, the pressure forces overwhelm any small normal-stress driven secondary flows in the inlet plane to given maximum values of around 0.07 of the bulk velocity in the exit plane. The precision of manufacture of the ducts means that the geometric boundary conditions are accurately known and the results are therefore suitable for calculation purposes. Similarly, the instrumentation allows the flow boundary conditions to be accurately determined.

Taylor, A. M. K. P.↗
NASA NTRS2006

Understanding Columbia's Reentry Problems

Soon after the Space Shuttle Columbia accident occurred last year, a group of CFD analysts from NASA centers and private industry was organized to help determine the cause of the accident. This group was under the direction of the Applied Aeroscience and CFD Branch of the Aeroscience and Flight Mechanics Division at the Johnson Space Center. For external flow simulations, noncommercia2 CFD codes that specialize in hypersonic or high Mach number flows were used. These tools were used to determine heating rates, pressures, and temperatures for a large number of vehicle damage scenarios. Lockheed Martin Space Operations was called upon to provide CFD support in the area of internal flows within the shuttle wing cavity, and for these simulations, FLUENT 6.1 was chosen. Two large-scale, simplified models were m to understand the flow patterns once a breach of the internal wing cavity was initiated. The results were primarily used to visualize flow patterns within the wing cavity. The first CFD model included the entire lee wing without the wheel well cavity. The purpose of the first model, which did not include the reinforced carbon-carbon (RCC) cavity along the wing leading edge, was to visualize the flow field within the wing cavity immediately after the leading edge spar breach, This model assumed that the flow coming into the wing cavity was normal to the spar. It included all of the primary vents that allow for flow between the main cavities of the wing. A six-inch diameter hole was modeled in the spar at the approximate location where the spar breach was judged to have occurred, which was between RCC panels 8 and 9. The results of the modeling showed that at this location, the high temperature, high velocity gas stream entering the wing cavity impinged on the outboard wheel well cavity. Instrumentation in the Shuttle wheel well cavity registered abnormal temperatures during reentry, so the FLUENT results helped support the conclusion of the accident investigation team that the spar breach was in the RCC panel 8-9 area, and that the initial spar breach was likely entering the wing cavity normal to the spar. This model also showed that the flow entering the wing cavity tended to swirl within the cavity just outboard of the wheel well, and did not initially penetrate further into the rear cavities of the wing. The second CFD model was a 2-D simulation of the left wing cavity and the RCC cavity. It was used to visualize the flow through the RCC breach, through the wing spar breach, and into the wing cavity directly outboard of the wheel well. The purpose of this model was to verify whether or not it was possible for the flow to come into the wing cavity normal to the leading edge spar. The results from this 2-D model showed that the internal structure behind the RCC panels probably deflected the flow entering the RCC cavity so that it impinged normal to the spar. As in the 3-D model, this deflected flow stream was Found to impinge on the wheel well outer wall. This model again supported the conclusions regarding the location of the spar breach and how the flow behaved inside the wing cavity.

Paul, Thomas↗
NASA NTRS1979

Evaluation of a new concept for reducing free-stream turbulence in wind tunnels

The feasibility of using a 45 deg-honeycomb for improving the flow quality in wind tunnels was investigated. The results of the experiments indicate that the turbulence levels in a wind tunnel using the 45 deg-honeycomb would be comparable to those obtainable using a conventional honeycomb, i.e., a honeycomb which is positioned normal to the mean flow in the settling chamber. However, this is true only when a 45 deg-screen is mounted immediately downstream of the honeycomb, and when some distance is provided between the trailing edges of the corner turning vanes and the upstream side of the 45 deg-honeycomb. This distance is required for adequate decay of the turning-vane wakes, thereby providing a reasonably uniform mean flow entering the honeycomb. Results demonstrate that this distance should be at least twice the spacing between the turning vanes. Even though the resulting turbulence intensity is the same downstream of both the 45 deg-honeycomb and the conventional honeycomb, there would have to be a significant improvement in the test-section turbulence intensity using the 45 deg-honeycomb to justify the additional expense in fabricating and installing this type of honeycomb.

Wigeland, R. A.↗
NASA NTRS1980

Nonparallel stability of three-dimensional compressible boundary layers. Part 1: Stability analysis

A compressible linear stability theory is presented for nonparallel three-dimensional boundary-layer flows, taking into account the normal velocity component as well as the streamwise and spanwise variations of the basic flow. The method of multiple scales is used to account for the nonparallelism of the basic flow, and equations are derived for the spatial evolution of the disturbance amplitude and wavenumber. The numerical procedure for obtaining the solution of the nonparallel problem is outlined.

El-Hady, N. M.↗
NASA NTRS1947

The Streamline Pattern in the Vicinity of an Oblique Airfoil

A method for determining the streamwise flow pattern of a nonviscous incompressible fluid about an oblique airfoil from the corresponding flow pattern about the airfoil in normal position is presented and illustrated in two examples. The method can be extended to account approximately for compressibility effects by applying the Prandtl-Glauert correction factor to the flow pattern that is normal to the leading edge of the airfoil. The method is expected to be useful in determining the shape of a fuselage or nacelle having a minimum of interference with the flow over a swept-back wing.

Watkins, Charles E↗
NASA NTRS1994

Variations of a global constraint factor in cracked bodies under tension and bending loads

Elastic-plastic finite-element analyses were used to calculate stresses and displacements around a crack in finite-thickness plates for an elastic-perfectly plastic material. Middle- and edge-crack specimens were analyzed under tension and bending loads. Specimens were 1.25 to 20 mm thick with various widths and crack lengths. A global constraint factor alpha(sub g), an averaged normal-stress to flow-stress ratio over the plastic region, was defined to simulate three-dimensional (3D) effects in two-dimensional (2D) models. For crack lengths and uncracked ligament lengths greater than four times the thickness, the global constraint factor was found to be nearly a unique function of a normalized stress-intensity factor (related to plastic-zone size to thickness ratio) from small- to large-scale yielding conditions for various specimen types and thickness. For crack length-to-thickness ratios less than four, the global constraint factor was specimen type, crack length and thickness dependent. Using a 2D strip-yield model and the global constraint factors, plastic-zone sizes and crack-tip displacements agreed reasonably well with the 3D analyses. For a thin sheet aluminum alloy, the critical crack-tip-opening angle during stable tearing was found to be independent of specimen type and crack length for crack length-to-thickness ratios greater than 4.

Newman, J. C., Jr.↗