Engineering Papers⌕ Search

SEARCH · Engineering Papers

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 235 records · Page 13

Rotor design of high tip speed low loading transonic fan.

This paper describes the design concepts, principles and details of a high tip speed transonic rotor having low aerodynamic loading. The purpose of the NASA sponsored investigation was to determine whether good efficiency and large stall margin could be obtained by designing a rotor to avoid flow separation associated with strong normal shocks. Fully supersonic flow through the outboard region of the rotor with compression accomplished by weak oblique shocks were major design concepts employed. Computer programs were written and used to derive blade sections consistent from the all-supersonic tip region to the all-subsonic hub region. Preliminary test results indicate attainment of design pressure ratio and design flow at design speed with about a 1.6 point decrement in efficiency and large stall margin.

Erwin, J. R.↗

Overview of NASA's microgravity combustion science and fire safety program

The study of fundamental combustion processes in a microgravity environment is a relatively new scientific endeavor. A few simple, precursor experiments were conducted in the early 1970's. Today the advent of the U.S. space shuttle and the anticipation of the Space Station Freedom provide for scientists and engineers a special opportunity -- in the form of long duration microgravity laboratories -- and need -- in the form of spacecraft fire safety and a variety of terrestrial applications -- to pursue fresh insight into the basic physics of combustion. Through microgravity, a new range of experiments can be performed since: (1) Buoyancy-induced flows are nearly eliminated; (2) Normally obscured forces and flows may be isolated; (3) Gravitational settling or sedimentation is nearly eliminated; and (4) Larger time or length scales in experiments become permissible.

Ross, Howard D.↗

Verification of thermal analysis codes for modeling solid rocket nozzles

One of the objectives of the Solid Propulsion Integrity Program (SPIP) at Marshall Space Flight Center (MSFC) is development of thermal analysis codes capable of accurately predicting the temperature field, pore pressure field and the surface recession experienced by decomposing polymers which are used as thermal barriers in solid rocket nozzles. The objective of this study is to provide means for verifications of thermal analysis codes developed for modeling of flow and heat transfer in solid rocket nozzles. In order to meet the stated objective, a test facility was designed and constructed for measurement of the transient temperature field in a sample composite subjected to a constant heat flux boundary condition. The heating was provided via a steel thin-foil with a thickness of 0.025 mm. The designed electrical circuit can provide a heating rate of 1800 W. The heater was sandwiched between two identical samples, and thus ensure equal power distribution between them. The samples were fitted with Type K thermocouples, and the exact location of the thermocouples were determined via X-rays. The experiments were modeled via a one-dimensional code (UT1D) as a conduction and phase change heat transfer process. Since the pyrolysis gas flow was in the direction normal to the heat flow, the numerical model could not account for the convection cooling effect of the pyrolysis gas flow. Therefore, the predicted values in the decomposition zone are considered to be an upper estimate of the temperature. From the analysis of the experimental and the numerical results the following are concluded: (1) The virgin and char specific heat data for FM 5055 as reported by SoRI can not be used to obtain any reasonable agreement between the measured temperatures and the predictions. However, use of virgin and char specific heat data given in Acurex report produced good agreement for most of the measured temperatures. (2) Constant heat flux heating process can produce a much higher heating rate than the radiative heating process. The results show that heating rates of 125 C/s (225 F/s) can be achieved. (3) The electrical resistance of the FM 5055 samples were about 150 Omega in the virgin state, and decreased to about 35 Omega when charred. A reliable scheme must be developed to electrically insulate the composite from the heater foil in order to prevent any current leakage into the sample which can result in volumetric heating of the char zone of the composite.

Keyhani, M.↗

Large-Scale Flow Structure in Turbulent Nonpremixed Flames under Normal- And Low-Gravity Conditions

It is well known that buoyancy has a major influence on the flow structure of turbulent nonpremixed jet flames. Buoyancy acts by inducing baroclinic torques, which generate large-scale vortical structures that can significantly modify the flow field. Furthermore, some suggest that buoyancy can substantially influence the large-scale structure of even nominally momentum-dominated flames, since the low velocity flow outside of the flame will be more susceptible to buoyancy effects. Even subtle buoyancy effects may be important because changes in the large-scale structure affects the local entrainment and fluctuating strain rate, and hence the structure of the flame. Previous studies that have compared the structure of normal- and micro-gravity nonpremixed jet flames note that flames in microgravity are longer and wider than in normal-gravity. This trend was observed for jet flames ranging from laminar to turbulent regimes. Furthermore, imaging of the flames has shown possible evidence of helical instabilities and disturbances starting from the base of the flame in microgravity. In contrast, these characteristics were not observed in normal-gravity. The objective of the present study is to further advance our knowledge of the effects of weak levels of buoyancy on the structure of transitional and turbulent nonpremixed jet flames. In later studies we will utilize the drop tower facilities at NASA Glenn Research Center (GRC), but the preliminary work described in this paper was conducted using the 1.25-second drop tower located at the University of Texas at Austin. A more detailed description of these experiments can be found in Idicheria et al.

Clemens, N. T.↗

Innovative Liner Concepts: Experiments and Impedance Modeling of Liners Including the Effect of Bias Flow

The study of normal impedance of perforated plate acoustic liners including the effect of bias flow was studied. Two impedance models were developed by modeling the internal flows of perforate orifices as infinite tubes with the inclusion of end corrections to handle finite length effects. These models assumed incompressible and compressible flows, respectively, between the far field and the perforate orifice. The incompressible model was used to predict impedance results for perforated plates with percent open areas ranging from 5% to 15%. The predicted resistance results showed better agreement with experiments for the higher percent open area samples. The agreement also tended to deteriorate as bias flow was increased. For perforated plates with percent open areas ranging from 1% to 5%, the compressible model was used to predict impedance results. The model predictions were closer to the experimental resistance results for the 2% to 3% open area samples. The predictions tended to deteriorate as bias flow was increased. The reactance results were well predicted by the models for the higher percent open area, but deteriorated as the percent open area was lowered (5%) and bias flow was increased. A fit was done on the incompressible model to the experimental database. The fit was performed using an optimization routine that found the optimal set of multiplication coefficients to the non-dimensional groups that minimized the least squares slope error between predictions and experiments. The result of the fit indicated that terms not associated with bias flow required a greater degree of correction than the terms associated with the bias flow. This model improved agreement with experiments by nearly 15% for the low percent open area (5%) samples when compared to the unfitted model. The fitted model and the unfitted model performed equally well for the higher percent open area (10% and 15%).

Kelly, Jeff↗

Stability aspects of diverging subsonic flow

The paper describes the computation of two-dimensional, subsonic, diverging internal flows and how they differ from the corresponding converging flows. Such diverging or decelerating flows occur in such obvious places as subsonic diffusers and inlets; however, such flows also occur in supersonic nozzles in the presence of a normal shock. The flow instability and its relation to the numerical method used, boundary conditions, and viscous effects are assessed both analytically and numerically. The inviscid flow is shown to be physically unstable and a poor representation of the true viscous flow.

Cline, M. C.↗

Nonlinear Reynolds stress model for turbulent shear flows

A nonlinear algebraic Reynolds stress model, derived using the renormalization group, is applied to equilibrium homogeneous shear flow and fully developed flow in a square duct. The model, which is quadratically nonlinear in the velocity gradients, successfully captures the large-scale inhomogeneity and anisotropy of the flows studied. The ratios of normal stresses, as well as the actual magnitudes of the stresses are correctly predicted for equilibrium homogeneous shear flow. Reynolds normal stress anisotropy and attendant turbulence driven secondary flow are predicted for a square duct. Profiles of mean velocity and normal stresses are in good agreement with measurements. Very close to walls, agreement with measurements diminishes. The model has the benefit of containing no arbitrary constants; all values are determined directly from the theory. It seems that near wall behavior is influenced by more than the large scale anisotropy accommodated in the current model. More accurate near wall calculations may well require a model for anisotropic dissipation.

Barton, J. Michael↗

Self-regulated cooling flows in elliptical galaxies and in cluster cores - Is exclusively low mass star formation really necessary?

A self-consistent treatment of the heating by supernovae associated with star formation in a spherically symmetric cooling flow in a cluster core or elliptical galaxy is presented. An initial stellar mass function similar to that in the solar neighborhood is adopted. Inferred star-formation rates, within the cooling region - typically the inner 100 kpc around dominant galaxies at the centers of cooling flows in XD clusters - are reduced by about a factor of 2, relative to rates inferred when the heat input from star formation is ignored. Truncated initial mass functions (IMFs) are also considered, in which massive star formation is suppressed in accordance with previous treatments, and colors are predicted for star formation in cooling flows associated with central dominant elliptical galaxies and with isolated elliptical galaxies surrounded by gaseous coronae. The low inferred cooling-flow rates around isolated elliptical galaxies are found to be insensitive to the upper mass cutoff in the IMF, provided that the upper mass cutoff exceeds 2 M solar mass. Comparison with observed colors favors a cutoff in the IMF above 1 M solar mass in at least two well-studied cluster cooling flows, but a normal IMF cannot be excluded definitively. Models for NGC 1275 support a young (less than about 3 Gyr) cooling flow. As for the isolated elliptical galaxies, the spread in colors is consistent with a normal IMF. A definitive test of the IMF arising via star formation in cooling flows requires either UV spectral data or supernova searches in the cooling-flow-centered galaxies.

Silk, J.↗

Further developments in simple total energy sensors

A simple total energy probe concept using principles of laminar flow around a cylinder and pressure distribution is reviewed and flow fields around cylinders normal to and inclined to the flow are described. A variety of bent-up probe configurations were tested to explore variations in geometry. Test results are presented on the effects of sensor length, hole location, and angle of sweep. Comparisons are made with other probe tests reported in the literature. Damping restrictors and their use in filtering gusts are summarized. Flow field effects, indicating the variables involved for different mounting locations on aircraft, are discussed.

Nicks, O. W.↗

The X-ray emission of normal elliptical galaxies - Steady state cooling flow models

Steady state, spherical, homogeneous cooling flow models are compared to the observed X-ray luminosities, spectra, and surface-brightness profiles of elliptical galaxies. Galaxies without heavy halos and SNe have temperatures which are too low and X-ray surface brightness profiles which are too rapidly decreasing with radius to fit the observations. Galaxies with heavy halos and SNe have luminosities which are too large except for the highest luminosity ellipticals. Galaxies without heavy halos but with SNe have temperatures which are probably too cool in the outer parts. Galaxies with heavy halos but without SNe, where the gas is heated by the motions of gas-losing stars and by infall and compression in a heavy halo potential, are favored. This suggests that elliptical galaxies have heavy halos of optically dark material.

Sarazin, Craig L.↗

A Detailed Investigation of Staged Normal Injection into a Mach 2 Flow

A study of the staged injection of two jets of air behind a rearward facing step into a Mach 2 flow was performed using the SPARK 3-D Navier-Stokes code. Calculated mole fraction distributions were compared with an extensive set of planar mole fraction measurements made with a laser induced iodine fluorescence technique. A statistical measure, the standard deviation, was used to help assess agreement between calculation and experiment. Overall, good agreement was found between calculated and measured values. Generally, agreement was better in the far field of the injectors. The effect of grid resolution was investigated by calculating solutions on grids of 60,000, 200,000, and 450,000 points. Differences in the solutions on the two finer grids were small. However, the mole fraction distributions were distinguishable. The effect of turbulence modeling was investigated by employing three different algebraic models for the jet turbulence: the Baldwin-Lomax model, the Prandtl mixing length model, and the Eggers mixing length model. Overall, the Eggers mixing length model was found to be superior for this case. Finally, the effect of the jet exit conditions was examined. A recently proposed Mach number distribution at the jet exit was found to slightly improve agreement between measurement and calculation.

Eklund, Dean R.↗

Shock turbulence interaction in the presence of mean shear - An application of rapid distortion theory

The present examination of the response of anisotropic turbulent flows to a shock wave uses incompressible, homogeneous rapid distortion theory and idealizes the shock-wave as 1D compression. Attention is given to the shock's effect on axisymmetric flow and both the normal and oblique cases for shear flow. The oblique angle between the directions of compression and shear significantly affects turbulence evolution. Compression amplifies all components of turbulence intensity.

Mahesh, Krishnan↗

Microgravity Combustion Science: 1995 Program Update

Microgravity greatly benefits the study of fundamental combustion processes. In this environment, buoyancy-induced flow is nearly eliminated, weak or normally obscured forces and flows can be isolated, gravitational settling or sedimentation is nearly eliminated, and temporal and spatial scales can be expanded. This document reviews the state of knowledge in microgravity combustion science with the emphasis on NASA-sponsored developments in the current period of 1992 to early 1995. The subjects cover basic research in gaseous premixed and diffusion-flame systems, flame structure and sooting, liquid droplets and pools, and solid-surface ignition and flame spread. They also cover applied research in combustion synthesis of ceramic-metal composites, advanced diagnostic instrumentation, and on-orbit fire safety. The review promotes continuing research by describing the opportunities for Principal Investigator participation through the NASA Research Announcement program and the available NASA Lewis Research Center ground-based facilities and spaceflight accommodations. This review is compiled by the members and associates of the NASA Lewis Microgravity Combustion Branch, and it serves as an update of two previous overview reports.

Ross, Howard D.↗

Drag characteristics of unsteady, perturbed boundary layer flows

A series of time-dependent numerical computations have been performed for the flow of an unsteady boundary layer over an embedded cavity, which is aligned normal to the flow direction. The unsteady flow is the result of the imposition of a Stuart vortex, at the inflow boundary of the computational domain, onto an otherwise unperturbed two-dimensional laminar boundary layer flow. This produces a strongly vortical, but spatially monochromatic, motion at inflow which is allowed to evolve downstream through the time-dependent numerical solution of appropriate conservation equations. As the imposed vortex evolves over the solid boundary, a weaker induced vortical motion is produced and together these perturbing motions interact with the fluid motion in the embedded cavity. An analysis of the pertinent dynamic variables as well as the time-dependent drag characteristics for this type of non-planar geometry is performed; with particular emphasis on the relative contribution of both pressure drag and frictional drag at various times in the evolution process.

Gatski, T. B.↗

Experimental study of laminar boundary layer receptivity to a traveling pressure field

Study is made of the production of Tollmien-Schlichting (T-S) waves within a Blasius boundary layer by an externally-imposed, two-dimensional, pressure field of a particular type. The field is that due to flow past a relatively small cylinder aligned parallel to the test plate and normal to the flow. The cylinder was carried on a circular path about an axis located well above the plate and such that the closest approach of the cylinder to the layer was several times the layer thickness or cylinder diameter. The turbulent wake did not, therefore, directly perturb the layer. Through adjustment of the circumferential velocity, the pressure field was swept along a certain region of the plate with controllable speed. Both the pressure field and the layer response are described in detail. It is shown that the production of T-S waves was sharply maximal for the case of circumferential speed near that of freely-propagating T-S waves, and that the coupling efficiency was extremely high for this case of resonance. The relevance of this energy transfer process to T-S wave production by the unsteady pressure fields resulting from wind tunnel wall turbulent boundary layers for both subsonic and supersonic flow is mentioned.

Kendall, James M.↗