Search NASA⌕ Search

SEARCH · Search NASA

Results for “Edge Computing”

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 901 records · Page 50

Two-dimensional hygrothermal diffusions into a finite width composite laminate

Two dimensional hygroscopic diffusion into composite laminates with free edges was obtained through the use of a second order finite difference scheme. A computer program was developed for the transient nonlinear coupled hygrothermal diffusion into finite width laminates with varying surface conditions along two edges. The formulation permits the diffusion coefficient to be a function of temperature, moisture concentration, and fiber orientation. The moisture distributions thus obtained are necessary for analysis of the moisture induced interlaminar stresses in the vicinity of free edges. It is shown that large diffusion times tend to eliminate the significance of stacking sequence, but gradients within and between layers are significant for all diffusion times.

Farley, G. L.↗

Design and implementation of a parallel unstructured Euler solver using software primitives

This paper is concerned with the implementation of a three-dimensional unstructured-grid Euler solver on massively parallel distributed-memory computer architectures. The goal is to minimize solution time by achieving high computational rates with a numerically efficient algorithm. An unstructured multigrid algorithm with an edge-based data structure has been adopted, and a number of optimizations have been devised and implemented to accelerate the parallel computational rates. The implementation is carried out by creating a set of software tools, which provide an interface between the parallelization issues and the sequential code, while providing a basis for future automatic run-time compilation support. Large practical unstructured grid problems are solved on the Intel iPSC/860 hypercube and Intel Touchstone Delta machine. The quantitative effects of the various optimizations are demonstrated, and we show that the combined effect of these optimizations leads to roughly a factor of 3 performance improvement. The overall solution efficiency is compared with that obtained on the Cray Y-MP vector supercomputer.

Das, R.↗

Computational Investigation of a Semi-Span Flap

The current study computationally examines one of the principle three-dimensional features of the flow over a high-lift system, the flow associated with a flap edge. Structured, overset grids were used in conjunction with an incompressible Navier-Stokes solver to compute the flow over a two-element high-lift configuration. The computations were run in a fully turbulent mode using the one-equation Baldwin-Barth model. Specific interest was given to the details of the flow in the vicinity of the flap edge, so the geometry was simplified to isolate this region. The geometry consisted of an unswept wing, which spanned a wind tunnel test section, equipped with a single element flap. Two flap configurations were computed; a full-span and a half-span Fowler flap. The chord based Reynolds number was 3.7 million for all cases. The results for the full-span flap agreed with two-dimensional experimental results and verified the method. Grid topologies and related issues for the half-span flap geometry are discussed. Results of the half-span flap case are presented with emphasis on the flow features associated with the flap edge.

Mathias, Donovan, L.↗

Discharge Coefficients for Axisymmetric Supersonic Nozzles

Computational Fluid Dynamics (CFD) analysis was used to compute effective nozzle discharge coefficients for subscale sharp-edged converging/diverging nozzles, with a variety of convergence half-angles, motor operating conditions, and two propellants with different ballistics. Convergence half-angles ranged from 0 to 80 deg. Analysis was conducted at total temperatures from 2946K (5303R) to 3346K (6023R) and over total pressures ranged from 2.72 MPa (395 psia) to 20.68 MPa (3000 psia). Area ratios (A(sub e)/A*) ranged from 7.43 to 9.39. Ratio of specific heats (gamma) ranged from 1.13 to 1.18. Throat and exit Reynolds numbers were calculated to be 8.26 x 10(exp 5) and 5.51 x 10(exp 5), respectively. Present results of nozzle discharge coefficients are reported and correlated as a function of nozzle convergence half-angle (theta(sub c)) and area ratios (A(sub e)/A*) for a constant divergence half-angle (theta(sub d)) of 15 deg. Computed discharge coefficients ranged from 0.88 to 0.97. They are compared with theory and experimental data available in literature. Available turbulence models with respect to grid refinements and heat transfer are discussed.

Ahmad, Rashid A.↗

In-flight pressure distributions and skin-friction measurements on a subsonic transport high-lift wing section

Flight experiments are being conducted as part of a multiphased subsonic transport high-lift research program for correlation with wind-tunnel and computational results. The NASA Langley Transport Systems Research Vehicle (B737-100 aircraft) is used to obtain in-flight flow characteristics at full-scale Reynolds numbers to contribute to the understanding of 3-D high-lift, multi-element flows including attachment-line transition and relaminarization, confluent boundary-layer development, and flow separation characteristics. Flight test results of pressure distributions and skin friction measurements were obtained for a full-chord wing section including the slat, main-wing, and triple-slotted, Fowler flap elements. Test conditions included a range of flap deflections, chord Reynolds numbers (10 to 21 million), and Mach numbers (0.16 to 0.40). Pressure distributions were obtained at 144 chordwise locations of a wing section (53-percent wing span) using thin pressure belts over the slat, main-wing, and flap elements. Flow characteristics observed in the chordwise pressure distributions included leading-edge regions of high subsonic flows, leading-edge attachment-line locations, slat and main-wing cove-flow separation and reattachment, and trailing-edge flap separation. In addition to the pressure distributions, limited skin-friction measurements were made using Preston-tube probes. Preston-tube measurements on the slat upper surface suggested relaminarization of the turbulent flow introduced by the pressure belt on the slat leading-edge surface when the slat attachment line was laminar. Computational analysis of the in-flight pressure measurements using two-dimensional, viscous multielement methods modified with simple-sweep theory showed reasonable agreement. However, overprediction of the pressures on the flap elements suggests a need for better detailed measurements and improved modeling of confluent boundary layers as well as inclusion of three-dimensional viscous effects in the analysis.

Yip, Long P.↗

A parametric study of the leading edge attachment line for the F-16XL

A three-dimensional Navier-Stokes code is used to computationally simulate the flow about a modified F-16XL. Transition mechanisms (e.g. attachment line location and crossflow instability) near the swept wing leading edge are analyzed in detail. Flow visualization is used to study the influence of angle-of-attack on the aforementioned transition mechanisms. Validation of the code is accomplished by comparison of numerically generated surface pressures with that obtained by in-flight experiments.

Flores, Jolen↗

MICCO: An Enhanced Multi-GPU Scheduling Framework for Many-Body Correlation Functions

Calculation of many-body correlation functions is one of the critical kernels utilized in many scientific computing areas, especially in Lattice Quantum Chromodynamics (Lattice QCD). It is formalized as a sum of a large number of contraction terms each of which can be represented by a graph consisting of vertices describing quarks inside a hadron node and edges designating quark propagations at specific time intervals. Due to its computation- and memory-intensive nature, real-world physics systems (e.g., multi-meson or multi-baryon systems) explored by Lattice QCD prefer to leverage multi-GPUs. Different from general graph processing, many-body correlation function calculations show two specific features: a large number of computation-/data-intensive kernels and frequently repeated appearances of original and intermediate data. The former results in expensive memory operations such as tensor movements and evictions. The latter offers data reuse opportunities to mitigate the data-intensive nature of many-body correlation function calculations. However, existing graph-based multi-GPU schedulers cannot capture these data-centric features, thus resulting in a sub-optimal performance for many-body correlation function calculations. To address this issue, this paper presents a multi-GPU scheduling framework, MICCO, to accelerate contractions for correlation functions particularly by taking the data dimension (e.g., data reuse and data eviction) into account. This work first performs a comprehensive study on the interplay of data reuse and load balance, and designs two new concepts: local reuse pattern and reuse bound to study the opportunity of achieving the optimal trade-off between them. Based on this study, MICCO proposes a heuristic scheduling algorithm and a machine-learning-based regression model to generate the optimal setting of reuse bounds. Specifically, MICCO is integrated into a real-world Lattice QCD system, Redstar, for the first time running on multiple GPUs. The evaluation demonstrates MICCO outperforms other state-of-art works, achieving up to 2.25× speedup in synthesized datasets, and 1.49× speedup in real-world correlation functions.

Wang, Qihan↗

Modern Design of Resonant Edge-Slot Array Antennas

Resonant edge-slot (slotted-waveguide) array antennas can now be designed very accurately following a modern computational approach like that followed for some other microwave components. This modern approach makes it possible to design superior antennas at lower cost than was previously possible. Heretofore, the physical and engineering knowledge of resonant edge-slot array antennas had remained immature since they were introduced during World War II. This is because despite their mechanical simplicity, high reliability, and potential for operation with high efficiency, the electromagnetic behavior of resonant edge-slot antennas is very complex. Because engineering design formulas and curves for such antennas are not available in the open literature, designers have been forced to implement iterative processes of fabricating and testing multiple prototypes to derive design databases, each unique for a specific combination of operating frequency and set of waveguide tube dimensions. The expensive, time-consuming nature of these processes has inhibited the use of resonant edge-slot antennas. The present modern approach reduces costs by making it unnecessary to build and test multiple prototypes. As an additional benefit, this approach affords a capability to design an array of slots having different dimensions to taper the antenna illumination to reduce the amplitudes of unwanted side lobes. The heart of the modern approach is the use of the latest commercially available microwave-design software, which implements finite-element models of electromagnetic fields in and around waveguides, antenna elements, and similar components. Instead of building and testing prototypes, one builds a database and constructs design curves from the results of computational simulations for sets of design parameters. The figure shows a resonant edge-slot antenna designed following this approach. Intended for use as part of a radiometer operating at a frequency of 10.7 GHz, this antenna was fabricated from dimensions defined exclusively by results of computational simulations. The final design was found to be well optimized and to yield performance exceeding that initially required.

Gosselin, R. B.↗

Emerging applications: Neuromorphic computing and reservoir computing

The emergence of doped hafnium oxide (HfO 2 )-based ferroelectric films has enabled highly scalable and silicon-compatible ferroelectric devices, opening new frontiers in neuromorphic and reservoir computing. Among these, ferroelectric field-effect transistors (FeFETs) are particularly promising due to their analog memory characteristics and unique polarization dynamics. These properties make FeFETs ideal candidates for artificial synapses in neuromorphic architectures, supporting deep neural networks and spiking neural networks based on leaky-integrate-and-fire (LIF) mechanisms. Beyond neuromorphic computing, FeFETs also play a crucial role in physical reservoir computing, leveraging their intrinsic nonlinear and history-dependent behavior for efficient real-time learning. This approach offers significant advantages for time-series processing and edge artificial intelligence (AI) applications, addressing the growing need for energy-efficient computing. As a result, this article explores the principles, key demonstrations, and future potential of FeFET-based neuromorphic and reservoir computing, highlighting their impact on next-generation AI hardware.

36 MATERIALS SCIENCE↗

Extended Simulations of Graphene Growth with Updated Rate Coefficients

New simulations of graphene growth in flame environments are presented. The simulations employ a kinetic Monte Carlo (KMC) algorithm coupled to molecular mechanics (MM) geometry optimization to track individual graphenic species as they evolve. Focus is given to incorporation of five-member rings and resulting curvature and edge defects. The model code has been re-written to be more computationally efficient enabling a larger set of simulations to be run, decreasing stochastic fluctuations in the averaged results. The model also includes updated rate coefficients for graphene edge reactions recently published in the literature. The new simulations are compared to results from the previous model as well as to hydrogen to carbon ratios recorded in experiment and calculated with alternate models.

30 DIRECT ENERGY CONVERSION↗

Spectroscopic, Computational, and Kinetic Studies of the Mu-Sulfide-Bridged Tetranuclear CuZ Cluster in N(2)O Reductase: PH Effect on the Edge Ligand and its Contribution to Reactivity

A combination of spectroscopy and density functional theory (DFT) calculations has been used to evaluate the pH effect at the Cu{sub Z} site in Pseudomonas nautica (Pn) nitrous oxide reductase (N{sub 2}OR) and Achromobacter cycloclastes (Ac) N{sub 2}OR and its relevance to catalysis. Absorption, magnetic circular dichroism, and electron paramagnetic resonance with sulfur K-edge X-ray absorption spectra of the enzymes at high and low pH show minor changes. However, resonance Raman (rR) spectroscopy of PnN{sub 2}OR at high pH shows that the 415 cm{sup -1} Cu-S vibration (observed at low pH) shifts to higher frequency, loses intensity, and obtains a 9 cm{sup -1} {sup 18}O shift, implying significant Cu-O character, demonstrating the presence of a OH{sup -} ligand at the Cu{sub I}Cu{sub IV} edge. From DFT calculations, protonation of either the OH{sup -} to H{sub 2}O or the {mu}{sub 4}-S{sup 2-} to {mu}{sub 4}-SH{sup -} would produce large spectral changes which are not observed. Alternatively, DFT calculations including a lysine residue at an H-bonding distance from the Cu{sub I}Cu{sub IV} edge ligand show that the position of the OH{sup -} ligand depends on the protonation state of the lysine. This would change the coupling of the Cu-(OH) stretch with the Cu-S stretch, as observed in the rR spectrum. Thus, the observed pH effect (pK{sub a} {approx} 9.2) likely reflects protonation equilibrium of the lysine residue, which would both raise E{sup o} and provide a proton for lowering the barrier for the N-O cleavage and for reduction of the [Cu{sub 4}S(im){sub 7}OH]{sup 2+} to the fully reduced 4Cu{sup I} active form for turnover.

59 BASIC BIOLOGICAL SCIENCES↗

A Locally Adapting Technique for Edge Detection using Image Segmentation

Rapid growth in the field of quantitative digital image analysis is paving the way for scientific researchers to make precise measurements about objects in an image. To compute quantities from an image such as the density of compressed materials or the velocity of a shockwave, object boundaries must first be determined. Images containing regions that each have a spatial trend in intensity are of particular interest here. For edge detection, we present a supervised, statistical image segmentation method that incorporates spatial information to locate boundaries between regions with overlapping intensity histograms, specifically for images where the regions are known but precise boundary locations are unknown. The segmentation of a pixel is determined by comparing its intensity to distributions from nearby pixel intensities, and a gradient of the segmented image indicates edge locations. Because of the statistical nature of the algorithm, we use maximum likelihood estimation to quantify uncertainty about each boundary. We demonstrate the success of this algorithm at locating boundaries and providing uncertainty bands on a radiograph of a multicomponent cylinder and on an optical image of a laser-induced shockwave.

97 MATHEMATICS AND COMPUTING↗

The design and implementation of a parallel unstructured Euler solver using software primitives

This paper is concerned with the implementation of a 3D unstructured-grid Euler-solver on massively parallel distributed-memory computer architectures. The goal is to minimize solution time by achieving high computational rates with a numerically efficient algorithm. An unstructured multigrid algorithm with an edge-based data-structure has been adopted, and a number of optimizations have been devised and implemented in order to accelerate the parallel computational rates. The implementation is carried out by creating a set of software tools, which ease the implementation of computational problems on parallel architecture machines by relieving the user of the low-level machine specific issues. The quantitative effect of the various optimizations are demonstrated, and we show that the combined effect of these optimizations leads to roughly a factor of three performance improvement. The overall solution efficiency is compared with that obtained on the CRAY-YMP vector supercomputer.

Das, R.↗

Shock Location Dominated Transonic Flight Loads on the Active Aeroelastic Wing

During several Active Aeroelastic Wing research flights, the shadow of the over-wing shock could be observed because of natural lighting conditions. As the plane accelerated, the shock location moved aft, and as the shadow passed the aileron and trailing-edge flap hinge lines, their associated hinge moments were substantially affected. The observation of the dominant effect of shock location on aft control surface hinge moments led to this investigation. This report investigates the effect of over-wing shock location on wing loads through flight-measured data and analytical predictions. Wing-root and wing-fold bending moment and torque and leading- and trailing-edge hinge moments have been measured in flight using calibrated strain gages. These same loads have been predicted using a computational fluid dynamics code called the Euler Navier-Stokes Three Dimensional Aeroelastic Code. The computational fluid dynamics study was based on the elastically deformed shape estimated by a twist model, which in turn was derived from in-flight-measured wing deflections provided by a flight deflection measurement system. During level transonic flight, the shock location dominated the wing trailing-edge control surface hinge moments. The computational fluid dynamics analysis based on the shape provided by the flight deflection measurement system produced very similar results and substantially correlated with the measured loads data.

Lokos, William A.↗

Fusion Simulation Program Execution Plan

The overall science goal of the FSP is to develop predictive simulation capability for magnetically confined fusion plasmas at an unprecedented level of integration and fidelity. This will directly support and enable effective U.S. participation in research related to the International Thermonuclear Experimental Reactor (ITER) and the overall mission of delivering practical fusion energy. The FSP will address a rich set of scientific issues together with experimental programs, producing validated integrated physics results. This is very well aligned with the mission of the ITER Organization to coordinate with its members the integrated modeling and control of fusion plasmas, including benchmarking and validation activities. Initial FSP research will focus on two critical areas: 1) the plasma edge and 2) whole device modeling including disruption avoidance. The first of these problems involves the narrow plasma boundary layer and its complex interactions with the plasma core and the surrounding material wall. The second requires development of a computationally tractable, but comprehensive model that describes all equilibrium and dynamic processes at a sufficient level of detail to provide useful prediction of the temporal evolution of fusion plasma experiments. The initial driver for the whole device model (WDM) will be prediction and avoidance of discharge-terminating disruptions, especially at high performance, which are a critical impediment to successful operation of machines like ITER. If disruptions prove unable to be avoided, their associated dynamics and effects will be addressed in the next phase of the FSP. The FSP plan targets the needed modeling capabilities by developing Integrated Science Applications (ISAs) specific to their needs. The Pedestal-Boundary model will include boundary magnetic topology, cross-field transport of multi-species plasmas, parallel plasma transport, neutral transport, atomic physics and interactions with the plasma wall. It will address the origins and structure of the plasma electric field, rotation, the L-H transition, and the wide variety of pedestal relaxation mechanisms. The Whole Device Model will predict the entire discharge evolution given external actuators (i.e., magnets, power supplies, heating, current drive and fueling systems) and control strategies. Based on components operating over a range of physics fidelity, the WDM will model the plasma equilibrium, plasma sources, profile evolution, linear stability and nonlinear evolution toward a disruption (but not the full disruption dynamics). The plan assumes that, as the FSP matures and demonstrates success, the program will evolve and grow, enabling additional science problems to be addressed. The next set of integration opportunities could include: 1) Simulation of disruption dynamics and their effects; 2) Prediction of core profile including 3D effects, mesoscale dynamics and integration with the edge plasma; 3) Computation of non-thermal particle distributions, self-consistent with fusion, radio frequency (RF) and neutral beam injection (NBI) sources, magnetohydrodynamics (MHD) and short-wavelength turbulence.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

The effect of spin-orbit coupling in band structure and edge states of bilayer graphene

Topological insulators are predicted to be useful ranging from spintronics to quantum computation. Graphene was first predicted to be the precursor of topological insulator by Kane-Mele. They developed a Hamiltonian model to describe the gap opening in graphene. In this work, we investigate the band structure of bilayer grapheme and also its edge states by using this model with analytical approach. The results of our calculation show that the gap opening occurs at K and K’ point in bilayer graphene.In addition, a pair of gapless edge modes occurs both in the zigzag and arm-chair configurations are no longer exist. There are gap created at the edge even though thery are very small.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

A viscous starter solution for shuttle flow field computations

A time dependent numerical procedure that calculates the viscous flow in the shock layer around the shuttle nose has been developed. It has been applied to the free stream conditions that may be encountered in the shuttle flight trajectory and found to be efficient and in satisfactory agreement with other steady and unsteady techniques. Two features associated with the present formulation are responsible for its successful applications. The first is that the flow field computation is made within the shock layer; the second is that the mapping of the computational region places proportionately more mesh points in the vicinity of the wall than in the remaining computational region. The numerical results may be used for low altitude flight to determine the edge conditions for a thin boundary layer, to provide initial boundary layer profiles for downstream boundary layer calculations, and to start supersonic computations. For high altitude flight, it may be used to determine the complete flow field in the nose region and also provide a starter solution for subsequent supersonic flow computations.

Li, C. P.↗

A Novel Coarsening Method for Scalable and Efficient Mesh Generation

In this paper, we propose a novel mesh coarsening method called brick coarsening method. The proposed method can be used in conjunction with any graph partitioners and scales to very large meshes. This method reduces problem space by decomposing the original mesh into fixed-size blocks of nodes called bricks, layered in a similar way to conventional brick laying, and then assigning each node of the original mesh to appropriate brick. Our experiments indicate that the proposed method scales to very large meshes while allowing simple RCB partitioner to produce higher-quality partitions with significantly less edge cuts. Our results further indicate that the proposed brick-coarsening method allows more complicated partitioners like PT-Scotch to scale to very large problem size while still maintaining good partitioning performance with relatively good edge-cut metric. Graph partitioning is an important problem that has many scientific and engineering applications in such areas as VLSI design, scientific computing, and resource management. Given a graph G = (V,E), where V is the set of vertices and E is the set of edges, (k-way) graph partitioning problem is to partition the vertices of the graph (V) into k disjoint groups such that each group contains roughly equal number of vertices and the number of edges connecting vertices in different groups is minimized. Graph partitioning plays a key role in large scientific computing, especially in mesh-based computations, as it is used as a tool to minimize the volume of communication and to ensure well-balanced load across computing nodes. The impact of graph partitioning on the reduction of communication can be easily seen, for example, in different iterative methods to solve a sparse system of linear equation. Here, a graph partitioning technique is applied to the matrix, which is basically a graph in which each edge is a non-zero entry in the matrix, to allocate groups of vertices to processors in such a way that many of matrix-vector multiplication can be performed locally on each processor and hence to minimize communication. Furthermore, a good graph partitioning scheme ensures the equal amount of computation performed on each processor. Graph partitioning is a well known NP-complete problem, and thus the most commonly used graph partitioning algorithms employ some forms of heuristics. These algorithms vary in terms of their complexity, partition generation time, and the quality of partitions, and they tend to trade off these factors. A significant challenge we are currently facing at the Lawrence Livermore National Laboratory is how to partition very large meshes on massive-size distributed memory machines like IBM BlueGene/P, where scalability becomes a big issue. For example, we have found that the ParMetis, a very popular graph partitioning tool, can only scale to 16K processors. An ideal graph partitioning method on such an environment should be fast and scale to very large meshes, while producing high quality partitions. This is an extremely challenging task, as to scale to that level, the partitioning algorithm should be simple and be able to produce partitions that minimize inter-processor communications and balance the load imposed on the processors. Our goals in this work are two-fold: (1) To develop a new scalable graph partitioning method with good load balancing and communication reduction capability. (2) To study the performance of the proposed partitioning method on very large parallel machines using actual data sets and compare the performance to that of existing methods. The proposed method achieves the desired scalability by reducing the mesh size. For this, it coarsens an input mesh into a smaller size mesh by coalescing the vertices and edges of the original mesh into a set of mega-vertices and mega-edges. A new coarsening method called brick algorithm is developed in this research. In the brick algorithm, the zones in a given mesh are first grouped into fixed size blocks called bricks. These brick are then laid in a way similar to conventional brick laying technique, which reduces the number of neighboring blocks each block needs to communicate. Contributions of this research are as follows: (1) We have developed a novel method that scales to a really large problem size while producing high quality mesh partitions; (2) We measured the performance and scalability of the proposed method on a machine of massive size using a set of actual large complex data sets, where we have scaled to a mesh with 110 million zones using our method. To the best of our knowledge, this is the largest complex mesh that a partitioning method is successfully applied to; and (3) We have shown that proposed method can reduce the number of edge cuts by as much as 65%.

97 MATHEMATICS AND COMPUTING↗