Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Large scale”

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 1,513 records · Page 84
All sourcesNASA NTRSDOE OSTI
DOE OSTI2024

Increased Occurrence of Large–Scale Windthrows Across the Amazon Basin

Convective storms with strong downdrafts create windthrows: snapped and uprooted trees that locally alter the structure, composition, and carbon balance of forests. Comparing Landsat imagery from subsequent years, we documented temporal and spatial variation in the occurrence of large (≥30 ha) windthrows across the Amazon basin from 1985 to 2020. Over 33 individual years, we detected 3179 large windthrows. Windthrow density was greatest in the central and western Amazon regions, with ~33% of all events occurring in ~3% of the monitored area. Return intervals for large windthrows in the same location of these “hotspot” regions are centuries to millennia, while over the rest of the Amazon they are >10,000 years. Our data demonstrate a nearly 4–fold increase in windthrow number and affected area between 1985 (78 windthrows and 6,900 ha) and 2020 (264 events and 32,170 ha), with more events of >500 ha size since 1990. Such extremely large events (>500 ha up to 2,543 ha) are responsible for interannual variation in the overall median (84 ± 5.2 ha; ±95% CI) and mean (147 ± 13 ha) windthrow area, but we did not find significant temporal trends in the size distribution of windthrows with time. Our results document increased damage from convective storms over the past 40 years in the Amazon, filling a gap in temporal records for tropical regions. Our publicly accessible large windthrow database provides a valuable tool for exploring dynamic conditions leading to damaging storms and their ecological impact on Amazon forests.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2020

Toward Large-Scale Image Segmentation on Summit

Semantic segmentation of images is an important computer vision task that emerges in a variety of application domains such as medical imaging, robotic vision and autonomous vehicles to name a few. While these domain-specific image analysis tasks involve relatively small image sizes (~ 10 2 × 10 2 ), there are many applications that need to train machine learning models on image data with extents that are orders of magnitude larger (~10 4 × 10 4 ). Training deep neural network (DNN) models on large extent images is extremely memory-intensive and often exceeds the memory limitations of a single graphical processing unit, a hardware accelerator of choice for computer vision workloads. Here, an efficient, sample parallel approach to train U-Net models on large extent image data sets is presented. Its advantages and limitations are analyzed and near-linear strong-scaling speedup demonstrated on 256 nodes (1536 GPUs) of the Summit supercomputer. Using a single node of the Summit supercomputer, an early evaluation of a recently released model parallel framework called GPipe is demonstrated to deliver ~ 2X speedup in executing a U-Net model with an order of magnitude larger number of trainable parameters than reported before. Performance bottlenecks for pipelined training of U-Net models are identified and mitigation strategies to improve the speedups are discussed. Together, these results open up the possibility of combining both approaches into a unified scalable pipelined and data parallel algorithm to efficiently train U-Net models with very large receptive fields on data sets of ultra-large extent images.

Seal, Sudip↗
DOE OSTI2025

Towards large-scale quantum optimization solvers with few qubits

Quantum computers hold the promise of more efficient combinatorial optimization solvers, which could be game-changing for a broad range of applications. However, a bottleneck for materializing such advantages is that, in order to challenge classical algorithms in practice, mainstream approaches require a number of qubits prohibitively large for near-term hardware. Here we introduce a variational solver for MaxCut problems over $m={{\mathcal{O}}}({n}^{k})$ binary variables using only n qubits, with tunable k > 1. The number of parameters and circuit depth display mild linear and sublinear scalings in m , respectively. Moreover, we analytically prove that the specific qubit-efficient encoding brings in a super-polynomial mitigation of barren plateaus as a built-in feature. Altogether, this leads to high quantum-solver performances. For instance, for m = 7000, numerical simulations produce solutions competitive in quality with state-of-the-art classical solvers. In turn, for m = 2000, experiments with n = 17 trapped-ion qubits feature MaxCut approximation ratios estimated to be beyond the hardness threshold 0.941. Our findings offer an interesting heuristics for quantum-inspired solvers as well as a promising route towards solving commercially-relevant problems on near-term quantum devices.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
NASA NTRS1984

Application of augmented-Lagrangian methods in meteorology: Comparison of different conjugate-gradient codes for large-scale minimization

A Lagrange multiplier method using techniques developed by Bertsekas (1982) was applied to solving the problem of enforcing simultaneous conservation of the nonlinear integral invariants of the shallow water equations on a limited area domain. This application of nonlinear constrained optimization is of the large dimensional type and the conjugate gradient method was found to be the only computationally viable method for the unconstrained minimization. Several conjugate-gradient codes were tested and compared for increasing accuracy requirements. Robustness and computational efficiency were the principal criteria.

Navon, I. M.↗
NASA NTRS1993

Large-scale density structures in the outer heliosphere

The Plasma Science experiment on the Voyager 2 spacecraft has measured the solar wind density from 1 to 38 AU. Over this distance, the solar wind density decreases as the inverse square of the heliocentric distance. However, there are large variations in this density at a given radius. Such changes in density are the dominant cause of changes in the solar wind ram pressure in the outer heliosphere and can cause large perturbations in the location of the termination shock of the solar wind. Following a simple model suggested by Suess, we study the non-equilibrium, dynamic location of the termination shock as it responds to these pressure changes. The results of this study suggest that the termination shock is rarely if ever at its equilibrium distance and may depart from that distance by as much as 50 AU at times.

Belcher, J. W.↗
NASA NTRS1990

Noise transmission characteristics of a large scale composite fuselage model

Results from an experimental test undertaken to study the basic noise transmission characteristics of a realistic, built-up composite fuselage model are presented. The floor-equipped stiffened composite cylinder was exposed to a number of different exterior noise source configurations in a large anechoic chamber. These exterior source configurations included two point sources located in the same plane on opposite sides of the cylinder, a single point source and a propeller simulator. The results indicate that the interior source field is affected strongly by exterior noise source phasing. Sidewall treatment is seen to reduce the overall interior sound pressure levels and dampen dominant acoustic resonances so that other acoustic modes can affect interior noise distribution.

Beyer, Todd B.↗
NASA NTRS1960

Data from a Static-Thrust Investigation of Large-Scale General Research VTOL-STOL Model in Ground Effect

The model was tested at two different elevations with the wing pivot at 1.008 and 2.425 propeller diameters above the ground. The slipstream of the propellers was deflected by tilting the wing and propellers, by deflections of large-chord trailing-edge flaps, and by combinations of flap deflection and wing tilt. Tests were conducted over a range of propeller disk loadings from 7.41 to 29.70 pounds per square foot. Force data for the complete model and pressure distributions for the wing and flaps behind one propeller were recorded and are presented in tabular form without analysis.

Huston, Robert J.↗
DOE OSTI2022

Large-scale frictionless jamming with power-law particle size distributions

Due to significant computational expense, discrete element method simulations of jammed packings of size-dispersed spheres with size ratios greater than 1:10 have remained elusive, limiting the correspondence between simulations and real-world granular materials with large size dispersity. Here, invoking a recently developed neighbor binning algorithm, we generate mechanically stable jammed packings of frictionless spheres with power-law size distributions containing up to nearly 4 000 000 particles with size ratios up to 1:100. By systematically varying the width and exponent of the underlying power laws, we analyze the role of particle size distributions on the structure of jammed packings. The densest packings are obtained for size distributions that balance the relative abundance of large-large and small-small particle contacts. Although the proportion of rattler particles and mean coordination number strongly depend on the size distribution, the mean coordination of nonrattler particles attains the frictionless isostatic value of six in all cases. The size distribution of nonrattler particles that participate in the load-bearing network exhibits no dependence on the width of the total particle size distribution beyond a critical particle size for low-magnitude exponent power laws. This signifies that only particles with sizes greater than the critical particle size contribute to the mechanical stability. However, for high-magnitude exponent power laws, all particle sizes participate in the mechanical stability of the packing.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2020

Large Scale Tensor Factorization via Parallel Sketches

Tensor factorization methods have recently gained increased popularity. A key feature that renders tensors attractive is the ability to directly model multi-relational data. In this work, we propose ParaSketch, a parallel tensor factorization algorithm that enables massive parallelism, to deal with large tensors. The idea is to compress the large tensor into multiple small tensors, decompose each small tensor in parallel, and combine the results to reconstruct the desired latent factors. Prior art in this direction entails potentially very high complexity in the (Gaussian) compression and final combining stages. Adopting sketching matrices for compression, the proposed method enjoys a dramatic reduction in compression complexity, and features a much lighter combining step. Moreover, theoretical analysis shows that the compressed tensors inherit latent identifiability under mild conditions, hence establishing correctness of the overall approach. Numerical experiments corroborate the theory and demonstrate the effectiveness of the proposed algorithm.

block term decomposition↗
DOE OSTI2020

Automated pipeline framework for processing of large-scale building energy time series data

Commercial buildings account for one third of the total electricity consumption in the United States and a significant amount of this energy is wasted. Therefore, there is a need for “virtual” energy audits, to identify energy inefficiencies and their associated savings opportunities using methods that can be non-intrusive and automated for application to large populations of buildings. Here we demonstrate virtual energy audits applied to large populations of buildings’ time-series smart-meter data using a systematic approach and a fully automated Building Energy Analytics (BEA) Pipeline that unifies, cleans, stores and analyzes building energy datasets in a non-relational data warehouse for efficient insights and results. This BEA pipeline is based on a custom compute job scheduler for a high performance computing cluster to enable parallel processing of Slurm jobs. Within the analytics pipeline, we introduced a data qualification tool that enhances data quality by fixing common errors, while also detecting abnormalities in a building’s daily operation using hierarchical clustering. We analyze the HVAC scheduling of a population of 816 buildings, using this analytics pipeline, as part of a cross-sectional study. With our approach, this sample of 816 buildings is improved in data quality and is efficiently analyzed in 34 minutes, which is 85 times faster than the time taken by a sequential processing. The analytical results for the HVAC operational hours of these buildings show that among 10 building use types, food sales buildings with 17.75 hours of daily HVAC cooling operation are decent targets for HVAC savings. Overall, this analytics pipeline enables the identification of statistically significant results from population based studies of large numbers of building energy time-series datasets with robust results. These types of BEA studies can explore numerous factors impacting building energy efficiency and virtual building energy audits. This approach enables a new generation of data-driven buildings energy analysis at scale.

36 MATERIALS SCIENCE↗
DOE OSTI2023

A parallel hub-and-spoke system for large-scale scenario-based optimization under uncertainty

Practical solution of stochastic programming problems generally requires the use of parallel computing resources. Here, we describe the open source package mpi-sppy, in which efficient and scalable parallelization is a central feature. We report computational experiments that demonstrate the ability to solve very large stochastic programming problems - including mixed-integer variants - in minutes of wall clock time, efficiently leveraging significant parallel computing resources. We report results for the largest publicly available instances of stochastic mixed-integer unit commitment problems, solving to provably tight optimality gaps. In addition, we introduce a novel software architecture that facilitates combinations of methods for accelerating convergence that can be combined in plug-and-play manner. Finally, the mpi-sppy package is written in Python, leverages the widely used Pyomo (http://www.pyomo.org) library for modeling mathematical programs, builds on existing MPI implementations to ensure efficiency and scalability, and is available via http://github.com/Pyomo/mpi-sppy.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS1983

Optical measurements of a large-scale laboratory Beam Plasma Discharge (BPD)

Optical emission measurements provided information on the spatial distribution of energetic (primary and suprathermal) electrons producing the emissions; power dissipation during beam transit of the system; and modifications of the primary beam velocity distribution. The measurements were carried out for BPD's produced in a very large vacuum chamber using a 391.4 nm scanning photometer and total light (red sensitive) TV systems. Results for conditions Eb 500 to 2000 V, injection pitch angle 0 to 75 deg, and neutral density are presented. The typical power loss in ionization in transit of the 20 m pathlength during BPD is from 3 to 10% of the input power, an order of magnitude greater than attributable to collisional ionization alone. Differences between rocket observations and the laboratory results are noted.

Bernstein, W.↗
NASA NTRS1998

The Study of Clusters of Galaxies and Large Scale Structures

Many research projects have been initiated and completed under support of this program. The results are summarized below. The work on the ROSAT Deep Survey has been successfully completed. A number of interesting results have been established within this joint MPE, Cal Tech, JHU, ST ScI, ESO collaboration. First, a very large fraction, 70-80 percent, of the X-ray background has been directly resolved into point sources. We have derived a new log N-log S for X-ray sources and have measured a source density of 970 sources per square degree at a limiting flux level. Care was taken in these studies to accurately model and measure the effects of sources confusion. This was possible because of our observing strategy which included both deep PSPC and HRI observations. No evidence of a population of narrow emission line galaxies has been established but some evidence for the evolution of low luminosity AGN (Seyfert galaxies) has been reported. The work on the ROSAT All Sky Survey Northern Cluster Survey has been substantially concluded but the publication of the list has been held up by the need to analyze newly re-calibrated data. This should result in publication over the next year. During the past year we have submitted a paper to the Astrophysical Journal which utilized a sample of clusters originally selected from the ROSAT All-sky survey at redshifts greater than 0.3. This sample was studied with ASCA to determine temperature and luminosity.

Source record↗
NASA NTRS1980

Large-scale modulation of galactic cosmic rays and anomalous He observed at not less than 16 AU with Pioneer 10

High-speed solar wind streams issued from specific ranges of solar longitude or evolved into a single stream at not less than 16 AU during 1978. Shock waves associated with these streams cause a large decrease in cosmic ray intensity starting at 1 AU in April and appears progressively with time outward to 16 AU; a stepwise intensity decrease then occurs for both the galactic cosmic rays and the anomalous He component. The flare-accelerated nuclei show a new aspect of solar wave propagation in the outer solar system; at about 16 to 18 AU, the intensity builds up because of dispersion, reaches a flux level, and finally declines when the outward rate of escape becomes dominant.

Pyle, K. R.↗
NASA NTRS1986

The relationship of the large-scale solar field to the interplanetary magnetic field - What will Ulysses find?

Using photospheric magnetic field observations obtained at the Stanford Wilcox Solar Observatory, results from a potential field model for the present solar cycle are given, and qualitative predictions of the IMF that Ulysses may encounter are presented. Results indicate that the IMF consists of large regions of opposite polarity separated by a neutral sheet (NS) (extended to at least 50 deg) and a four-sector structure near solar minimum (produced by small quadripolar NS warps). The latitudinal extent of the NS increases following minimum and the structure near maximum includes multiple NSs, while a simplified IMF is found during the declining phase.

Hoeksema, J. T.↗
DOE OSTI2021

A large-scale comparison of Artificial Intelligence and Data Mining (AI&DM) techniques in simulating reservoir releases over the Upper Colorado Region

In recent years, the Artificial Intelligence and Data Mining (AI&DM) models have become popular tools in assisting various aspects of reservoir operation. However, the practical uses are still rarely reported. Comparison experiment of many AI&DM models over a large number of reservoir cases is particularly valuable to help reservoir operators first examine the usefulness and transferability of different AI&DM models, and then identify the most stable and reliable AI&DM model in assist of various decision-making processes. In this study, a total of 12 AI&DM models with different parameterizations and simulation scenarios are comprehensively tested out and compared in simulating the controlled reservoir outflows of 33 reservoir cases over the Upper Colorado Region, United States. Results show that the Random Forecast and the Long-Short-Term-Memory model could consistently derive the best statistical performance than other models under the baseline simulation scenario. The employed AI&DM models could obtain satisfactory statistical interquartile ranges (25–75%) between [0.6–0.9], [0.3–0.8], and [0.2–0.8], for CORR, NSE, and KGE measurements, respectively, and [1.5–6.5], [–15 to 20], and [0.5–8.5] for the normalized RMSE, PBIAS and RSR measurements, respectively. Results also show Multi-Layer Perceptron model and Extreme Gradient Boosting Tree Algorithm produced more stable and superior performance than other models under more complex input scenarios. We also found that the performance of different AI&DM models are closely relevant to the reservoir elevations, sizes, and functionalities. Discussions were made about the sensitivity of AI&DM models’ parameterizations and the key advantages of AI&DM models over the rule-based reservoir models. We further identify that the main advantage of AI&DM models is the flexibility in designing input structures, whereas the rule-based simulation model is rather limited. Future studies were suggested regarding the best way reservoir operators and researchers could use, select, and apply different AI&DM models in simulating reservoir releases under different natural and modeling environments. Finally, this comparison study also serves as a reference and a piece of groundwork for further promoting the practical uses of AI&DM models in assisting reservoir operation.

54 ENVIRONMENTAL SCIENCES↗
NASA NTRS2011

Simulations of Cyclone Sidr in the Bay of Bengal with a High-Resolution Model: Sensitivity to Large-Scale Boundary Forcing

The predictability of Cyclone Sidr in the Bay of Bengal was explored in terms of track and intensity using the Advanced Research Hurricane Weather Research Forecast (AHW) model. This constitutes the first application of the AHW over an area that lies outside the region of the North Atlantic for which this model was developed and tested. Several experiments were conducted to understand the possible contributing factors that affected Sidr s intensity and track simulation by varying the initial start time and domain size. Results show that Sidr s track was strongly controlled by the synoptic flow at the 500-hPa level, seen especially due to the strong mid-latitude westerly over north-central India. A 96-h forecast produced westerly winds over north-central India at the 500-hPa level that were notably weaker; this likely caused the modeled cyclone track to drift from the observed actual track. Reducing the model domain size reduced model error in the synoptic-scale winds at 500 hPa and produced an improved cyclone track. Specifically, the cyclone track appeared to be sensitive to the upstream synoptic flow, and was, therefore, sensitive to the location of the western boundary of the domain. However, cyclone intensity remained largely unaffected by this synoptic wind error at the 500-hPa level. Comparison of the high resolution, moving nested domain with a single coarser resolution domain showed little difference in tracks, but resulted in significantly different intensities. Experiments on the domain size with regard to the total precipitation simulated by the model showed that precipitation patterns and 10-m surface winds were also different. This was mainly due to the mid-latitude westerly flow across the west side of the model domain. The analysis also suggested that the total precipitation pattern and track was unchanged when the domain was extended toward the east, north, and south. Furthermore, this highlights our conclusion that Sidr was influenced from the west side of the domain. The displacement error was significantly reduced after the domain size from the western model boundary was decreased. Study results demonstrate the capability and need of a high-resolution mesoscale modeling framework for simulating the complex interactions that contribute to the formation of tropical cyclones over the Bay of Bengal region

Kumar, Anil↗
NASA NTRS1998

Cross Correlation of the 2-10 keV X-Ray Background with Radio Sources: Constraining the Large Scale Structure of the X-Ray Background

We present cross-correlation analyses of the HEAO 2-10 keV diffuse X-ray map with both the combined GB6/Parkes-MIT-NRAO (GB6-PMN) 5 GHz and the FIRST 1.4 GHz radio surveys. The cross-correlation functions (CCFs) of both radio surveys with the unresolved X-ray background were detected at the 5 u level. While the large angular resolution (3 deg) of the X-ray map makes it difficult to separate the contributions of clustering from those of Poisson fluctuations, the amplitude of the CCF provides important constraints on the X-ray emissivity of the radio sources, as well as on the clustering properties of radio and X-ray sources. These constraints are subject to a number of modeling parameters, e.g., X-ray luminosity evolution, clustering evolution, the radio luminosity function, cosmological model, etc. For reasonable choices of parameters the X-ray/FIRST CCF is consistent with a correlation scale length of 6/h Mpc. This is somewhat smaller than the scale length inferred from the autocorrelation function of the FIRST survey and implies that X-ray sources are less strongly clustered than strong radio sources, a result that is consistent with previous constraints on X-ray clustering. The X-ray/GB6-PMN CCF is several times larger and is likely to be dominated by Poisson fluctuations. This implies that approx. 2% of the diffuse X-ray background arises from the GB6-PMN sources.

Boughn, Stephen P.↗