Search NASA⌕ Search

SEARCH · Search NASA

Results for “Edge Modeling”

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 1,405 records · Page 78

All Optical Neural Networks for Low Power Edge Computing

We developed a simplistic physics-based model of an all-optical neural network that mimics the encoder part of an autoencoder neural network for image compression. Our approach relies on the generation of a MATLAB-based model for both data compression and decompression and utilizes MATLAB's built-in autoencoder networks in combination with simple propagation of optical fields between layers constituting phase elements via Fourier transform. We optimize the phase elements using the particle swarm optimization technique and using our model, we demonstrate a compression ratio of 25% for 2828-pixel input images containing numeric digits from 0 to 9.

97 MATHEMATICS AND COMPUTING↗

Development of a linear unsteady aerodynamic analysis for finite-deflection subsonic cascades

A linear unsteady potential flow analysis, which accounts for the effects of blade geometry and steady turning, is being developed to predict the aerodynamic response to blade vibrations in the fan, compressor or turbine stages of modern jet engines. In previous work numerical solutions were restricted to cascades of sharp-edged blades aligned with the mean flow. Under the present effort the solution procedure has been extended to treat blades with rounded or blunt edges. As part of this effort an analytical model-problem study has been conducted to clarify the behavior of first-order perturbation solutions in the vicinity of airfoil edges. Further, a numerical approximation using concepts from singular perturbation theory has been developed to resolve the unsteady boundary value problem for cascades of blunt leading-edged blades. Numerical results for NACA 0012 cascades, including detailed results in the vicinity of a blade leading edge, are presented and evaluated.

Verdon, J. M.↗

Laser transit anemometer measurements on a slender cone in the Langley unitary plan wind tunnel

A laser transit anemometer (LTA) system was used to probe the boundary layer on a slender (5 degree half angle) cone model in the Langley unitary plan wind tunnel. The anemometer system utilized a pair of laser beams with a diameter of 40 micrometers spaced 1230 micrometers apart to measure the transit times of ensembles of seeding particles using a cross-correlation technique. From these measurements, boundary layer profiles around the model were constructed and compared with CFD calculations. The measured boundary layer profiles representing the boundary layer velocity normalized to the edge velocity as a function of height above the model surface were collected with the model at zero angle of attack for four different flow conditions, and were collected in a vertical plane that bisected the model's longitudinal center line at a location 635 mm from the tip of the forebody cone. The results indicate an excellent ability of the LTA system to make velocity measurements deep into the boundary layer. However, because of disturbances in the flow field caused by onboard seeding, premature transition occurred implying that upstream seeding is mandatory if model flow field integrity is to be maintained. A description and results of the flow field surveys are presented.

Humphreys, William M., Jr.↗

A Transactive Approach for Service Restoration Utilizing Customer Load Flexibility and Grid-Edge Resources

This paper develops a transactive energy system model to restore electricity to customers in an isolated distribution system after an outage. The model engages a variety of customer types -- prosumers, flexible loads, critical/noncritical customers, and distributed generators -- as active participants in the restoration process. Unlike many existing transactive approaches, the proposed model is developed for service restoration and accounts for various customer types and their autonomy and privacy through an iterative approach to determine the optimal market price, while maintaining system-level power flow and voltage constraints. The advantages of the proposed approach are numerically validated on a modified IEEE 123-bus test system.

distributed energy resources↗

Investigation at Mach Numbers of 0.20 to 3.50 of Blended Wing-Body Combinations of Sonic Design with Diamond, Delta, and Arrow Plan Forms

The models had aspect-ratio-2 diamond, delta, and arrow wings with the leading edges swept 45.00 deg, 59.04 deg, and 70.82 deg, respectively. The wing sections were computed by varying the section shape along with the body radii (blending process) to match the prescribed area distribution and wing plan form. The wing sections had an average value of maximum thickness ratio of about 4 percent of the local chords in a streamwise direction. The models were tested with transition fixed at Reynolds numbers of about 4,000,000 to 9,000,0000, based on the mean aerodynamic chord of the wings. The effect of varying Reynolds number was checked at both subsonic and supersonic speeds. The diamond model was superior to the other plan forms at transonic speeds ((L/D)max = 11.00 to 9.52) because of its higher lift-curve slope and near optimum wave drag due to the blending process. For the wing thickness tested with the diamond model, the marked body and wing contouring required for transonic conditions resulted in a large wave-drag penalty at the higher supersonic Mach numbers where the leading and trailing edges of the wing were supersonic. Because of the low sweep of the trailing edge of the delta model, this configuration was less adaptable to the blending process. Removing a body bump prescribed by the Mach number 1.00 design resulted in a good supersonic design. This delta model with 10 percent less volume was superior to the other plan forms at Mach numbers of 1.55 to 2.35 ((L/D)max = 8.65 to 7.24), but it and the arrow model were equally good at Mach numbers of 2.50 to 3.50 ((L/D)max - 6.85 to O.39). At transonic speeds the arrow model was inferior because of the reduced lift-curve slope associated with its increased sweep and also because of the wing base drag. The wing base-drag coefficients of the arrow model based on the wing planform area decreased from a peak value of 0.0029 at Mach number 1.55 to 0.0003 at Mach number 3.50. Linear supersonic theory was satisfactory for predicting the aerodynamic trends at Mach numbers from 1.55 to 3.50 of lift-curve slope, wave drag, drag due to lift, aerodynamic-center location, and maximum lift-drag ratios for each of the models.

Holdaway, George H.↗

Measurements of Reynolds analogy for a hypersonic turbulent boundary layer on a nonadiabatic flat plate.

The purpose of the described experiment was to obtain simultaneous measurements of skin friction and heat transfer at hypersonic Mach numbers, especially with conditions of considerable heat transfer. The experiment was conducted in air in the Ames Hypersonic Wind Tunnel, in which cold air was passed through an alumina storage heater system and heated to total temperatures ranging from about 670 to 1170 K. The nozzle was contoured to produce a flow at Mach 7.4. Thin-skin heat-transfer gauges were placed at equal intervals along the centerline of a sharp-edge flat plate, which was used as the model. A skin-friction balance and a boundary-layer pitot-pressure rake were mounted on each side of the centerline at a certain distance from the leading edge.

Keener, E. R.↗

Numerical Analysis of Incipient Separation on 53 Deg Swept Diamond Wing

A systematic analysis of incipient separation and subsequent vortex formation from moderately swept blunt leading edges is presented for a 53 deg swept diamond wing. This work contributes to a collective body of knowledge generated within the NATO/STO AVT-183 Task Group titled 'Reliable Prediction of Separated Flow Onset and Progression for Air and Sea Vehicles'. The objective is to extract insights from the experimentally measured and numerically computed flow fields that might enable turbulence experts to further improve their models for predicting swept blunt leading-edge flow separation. Details of vortex formation are inferred from numerical solutions after establishing a good correlation of the global flow field and surface pressure distributions between wind tunnel measurements and computed flow solutions. From this, significant and sometimes surprising insights into the nature of incipient separation and part-span vortex formation are derived from the wealth of information available in the computational solutions.

Frink, Neal T.↗

Development and validation of a predictive model for the pedestal height

The pressure at the top of the edge transport barrier (or 'pedestal height') strongly impacts tokamak fusion performance. Predicting the pedestal height in future devices such as ITER [ITER Physics Basis Editors, Nucl. Fusion 39, 2137 (1999)] remains an important challenge. While uncertainties remain, magnetohydrodynamic stability calculations at intermediate wavelength (the ''peeling-ballooning'' model), accounting for diamagnetic stabilization, have been largely successful in determining the observed maximum pedestal height, when the edge barrier width is taken as an input. Here, we develop a second relation between the pedestal width in normalized poloidal flux ({delta}) and pedestal height ({delta}=0.076{beta}{sub {theta}}{sub ,ped}{sup 1/2}), using an argument based upon kinetic ballooning mode turbulence and observation. Combining this relation with direct calculations of peeling-ballooning stability yields two constraints, which together determine both the height and width of the pedestal. The resulting model, EPED1, allows quantitative prediction of the pedestal height and width in both existing and future experiments. EPED1 is successfully tested both against a dedicated experiment on the DIII-D [J. L. Luxon, Nucl. Fusion 42, 614 (2002)] tokamak, in which predictions were made before the experiment, and against a broader DIII-D data set, including ITER demonstration discharges. EPED1 is found to quantitatively capture the observed complex dependencies of the pedestal height and width. An initial set of pedestal predictions for the ITER device is presented.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Edge detection for synthetic aperture radar and other noisy images

The development is examined of a new edge detector which is shown to perform adequately in the non-Gaussian multiplicative noise environment which characterizes radar images. This edge detector operates over larger local neighborhood and is less susceptible to noise than previous edge detectors and is therefore more suitable for radar. In addition, a radar image noise model is employed for the design of this new operator. This edge detector is unique in that it is assumed that every local area belongs to either the class of local areas not containing edges or to the class of local areas containing edges. Each pixel's local neighborhood is then assigned to one of these two classes using a statistical hypothesis (a likelihood ratio) test. It is demonstrated that this algorithm is useful for detecting edges in radar images.

Frost, V. S.↗

Understanding important features of deep learning models for segmentation of high-resolution transmission electron microscopy images

Cutting edge deep learning techniques allow for image segmentation with great speed and accuracy. However, application to problems in materials science is often difficult since these complex models may have difficultly learning meaningful image features that would enable extension to new datasets. In situ electron microscopy provides a clear platform for utilizing automated image analysis. In this work, we consider the case of studying coarsening dynamics in supported nanoparticles, which is important for understanding, for example, the degradation of industrial catalysts. By systematically studying dataset preparation, neural network architecture, and accuracy evaluation, we describe important considerations in applying deep learning to physical applications, where generalizable and convincing models are required. With a focus on unique challenges that arise in high-resolution images, we propose methods for optimizing performance of image segmentation using convolutional neural networks, critically examining the application of complex deep learning models in favor of motivating intentional process design.

36 MATERIALS SCIENCE↗

A search for the iron absorption edge in the tail of an X-ray burst from X1636 - 53

Model atmosphere calculations of the spectrum of a neutron star cooling after an X-ray burst show that the photoelectric edge of iron should be prominent. No clear evidence for such a redshifted feature in the spectrum of a burst from X1636 - 53 is found, and it is concluded that the iron abundance there must be less than 0.3 solar. Unless the iron abundance of the surface matter on the neutron star is highly time-dependent, the present result argues against the 4.1-keV absorption line seen in some bursts from X1636 - 53 by Waki et al. (1984) being due to iron. The iron edge will be a powerful diagnostic of the surface redshift of the neutron star in burst sources where the iron abundance is more nearly solar.

Day, C. S. R.↗

Multiphysics Simulation of Active Hypersonic Lip Cooling

This article describes the application of the Multidisciplinary Analysis (MDA) solver, Spectrum, in analyzing a hydrogen-cooled hypersonic cowl leading-edge structure. Spectrum, a multiphysics simulation code based on the finite element method, addresses compressible and incompressible fluid flow, structural, and thermal modeling, as well as the interactions between these disciplines. Fluid-solid-thermal interactions in a hydrogen impingement-cooled leading edge are predicted using Spectrum. Two- and semi-three-dimensional models are considered for a leading edge impingement coolant, concept under either specified external heat flux or aerothermodynamic heating from a Mach 5 external flow interaction. The solution accuracy is demonstrated from mesh refinement analysis. With active cooling, the leading edge surface temperature is drastically reduced from 1807 K of the adiabatic condition to 418 K. The internal coolant temperature profile exhibits a sharp gradient near channel/solid interface. Results from two different cooling channel configurations are also presented to illustrate the different behavior of alternative active cooling schemes.

Melis, Matthew E.↗

Investigation of the selective sites on graphitic carbons for oxidative dehydrogenation of isobutane

Reaction network analysis of the oxidative dehydrogenation (ODH) reaction of isobutane over model carbon catalysts with tailored open-edge graphitic structure and uniform oxygenated functionalities was used to identify the selective pathways for the formation of isobutene. Carbon-based materials have been widely used in catalysis, but the active sites are not well-understood due to the complexity of the carbon structure. Correct identification of these sites is essential for learning how to manipulate material structure to achieve high catalytic yields of the desired products. In this study, we created model catalysts with controllable surface concentration of oxygenated functionalities based on graphitized mesoporous carbon (GMC). The concentration of the oxygenated functional groups on the surface of the GMCs was varied in a continuous manner through a simply control of the burn-off level. Our studies reveal that the ODH reaction of isobutane on carbon catalysts is a parallel-consecutive pathway with partial oxidative dehydrogenation for the formation of isobutene and deep oxidation pathway for the direct formation of CO and CO2 from isobutane. These two pathways show different dependence on the quinone-type oxygen sites: the rate constant leading to the desired partial oxidation product does not show a strong correlation to the density of oxygen sites, whereas the rate constant leading to the unselective COx products increases continuously with the density of oxygen sites. These results may give insight on the rational design of carbon catalysts for ODH reactions with desirable properties.

2-METHYLPROPANE↗

Community detection robustness of graph neural networks

Graph neural networks (GNNs) are increasingly widely used for community detection in attributed networks. They combine structural topology with node attributes through message passing and pooling. However, their robustness or lack thereof with respect to different perturbations and targeted attacks in conjunction with community detection tasks is not well understood. To shed light on latent mechanisms behind GNN sensitivity on community detection tasks, we conduct a systematic computational evaluation of six widely adopted GNN architectures graph convolutional network, graph attention network, graph sample and aggregate (GraphSAGE), differentiable pooling (DiffPool), minimum cut pooling (MinCUT), and deep modularity networks (DMoN). The analysis covers three perturbation categories: node attribute manipulations, edge topology distortions, and adversarial attacks. We use element-centric similarity as the evaluation metric on synthetic benchmarks and real-world citation networks. Our findings indicate that supervised GNNs tend to achieve higher baseline accuracy, while unsupervised methods, particularly DMoN, maintain stronger resilience under targeted and adversarial perturbations. Furthermore, robustness appears to be strongly influenced by community strength, with well-defined communities reducing performance loss. Across all models, node attribute perturbations associated with targeted edge deletions and shifts in attribute distributions tend to cause the largest degradation in community recovery. These findings highlight important trade-offs between accuracy and robustness in GNN-based community detection and offer insights into selecting architectures resilient to noise and adversarial attacks.

Goel, Jaidev [Virginia Polytechnic Inst. and State↗

Linear and Nonlinear Analyses of a Wind-Tunnel Balance

The NASA Langley Research Center (LaRC) has been designing strain-gauge balances for utilization in wind tunnels since its inception. The utilization of balances span a wide variety of aerodynamic tests. A force balance is an inherently critically stressed component due to the requirements of measurement sensitivity. Force balance stress analysis and acceptance criteria are under review due to LaRC wind tunnel operational safety requirements. This paper presents some of the analyses done at NASA LaRC. Research and analyses were performed in order to investigate the structural integrity of the balances and better understand their performance. The analyses presented in this paper are helpful in understanding the overall behavior of an existing balance and can also be used in design of new balances to enhance their performance. As a first step, maximum load combination is used for linear structural analysis. When nonlinear effects are encountered, the analysis is extended to include the nonlinearities. Balance 1621 is typical for LaRC designed balances and was chosen for this study due to its traditional high load capacity, Figure 1. Maximum loading occurs when all 6 components are applied simultaneously with their maximum value allowed (limit load). This circumstance normally will not occur in the wind tunnel. However, if it occurs, is the balance capable of handling the loads with an acceptable factor of safety? Preliminary analysis using Pro/Mechanica indicated that this balance might experience nonlinearity. It was decided to analyze this balance by using NASTRAN so that a nonlinear analysis could be conducted. Balance 1621 was modeled and meshed in PATRAN for analysis in NASTRAN. The model from PATRAN/NASTRAN is compared to the one from Pro/Mechanica. For a complete analysis, it is necessary to consider all the load cases as well as use a dense mesh near all the edges. Because of computer limitations, it is not feasible to analyze model with the dense mesh near all edges. In the present study, the dense mesh is limited to the surface on the end of the axial sections.

Karkehabadi, R.↗

Assessment and development of second order turbulence models

Second order turbulence models describe the effect of mean flow and external agencies (such as buoyancy) on the evolution of turbulence. Therefore, in principle, these models give a more accurate description of complicated flow fields (e.g. flows with large anisotropy in turbulence, such as near the leading edge of a turbine blade) than the two equation models. The objectives of this document are to assess the performance of the various second order turbulence models in benchmark flows and to seek improvements where necessary in models for the pressure correlation term in the scalar flux equation and for the scalar dissipation equation.

Shabbir, Aamir↗

KINEMATICS AT THE EDGE OF THE GALACTIC BULGE: EVIDENCE FOR CYLINDRICAL ROTATION

We present new results from BRAVA, a large-scale radial velocity survey of the Galactic bulge, using M giant stars selected from the Two Micron All Sky Survey catalog as targets for the Cerro Tololo Inter-American Observatory 4 m Hydra multi-object spectrograph. The purpose of this survey is to construct a new generation of self-consistent bar models that conform to these observations. We report the dynamics for fields at the edge of the Galactic bulge at latitudes b = -8 deg. and compare to the dynamics at b = -4 deg. We find that the rotation curve V(r) is the same at b = -8 deg. as at b = -4 deg. That is, the Galactic boxy bulge rotates cylindrically, as do boxy bulges of other galaxies. The summed line-of-sight velocity distribution at b = -8 deg. is Gaussian, and the binned longitude-velocity plot shows no evidence for either a (disk) population with cold dynamics or for a (classical bulge) population with hot dynamics. The observed kinematics are well modeled by an edge-on N-body bar, in agreement with published structural evidence. Our kinematic observations indicate that the Galactic bulge is a prototypical product of secular evolution in galaxy disks, in contrast with stellar population results that are most easily understood if major mergers were the dominant formation process.

79 ASTRONOMY AND ASTROPHYSICS↗