Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “gradient sampling”

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 199 records · Page 11
All sourcesNASA NTRSDOE OSTI
DOE OSTI2024

Validation of Hole-Drilling Residual Stress Measurements in Workpieces of Various Thickness

A recent revision to the ASTM E837 standard for near-surface residual stress measurement by the hole-drilling method describes a new thickness-dependent stress calculation procedure applicable to “thin” and “intermediate” workpieces for which strain versus depth response depends on workpiece thickness. This new calculation procedure differs from that of the prior standard, which applies only to thick workpieces with strain versus depth response independent of thickness. Herein we assess the new calculation procedures by performing hole-drilling residual stress measurements in samples with a range of thickness. Near-surface residual stress is measured in a thick aluminum plate containing near-surface residual stress from a uniform shot peening treatment, and in samples of different thickness removed from the plate at the peened surface. A finite element (FE) model is used to assess consistency between measured residual stress across the range of sample thickness. Measured residual stress varies with sample thickness, with thinner samples exhibiting smaller near-surface compressive stress and a larger gradient of subsurface stress. These trends are consistent with both observed bending (curvature) of the removed samples and the trend in FE-calculated expected residual stress. The measured and expected residual stresses are in good agreement for samples of intermediate thickness, but the agreement decreases with sample thickness. Measured residual stress is invariant with gage circle diameter. The new thickness-dependent stress calculation procedure for hole-drilling provides meaningful improvement compared to thick-workpiece calculations.

42 ENGINEERING↗
NASA NTRS1982

Experimental observation of the thermocapillary driven motion of bubbles in a molten glass under low gravity conditions

An experiment was carried out on board a Space Processing Applications Rocket with the aim of demonstrating bubble migration in molten glass due to a temperature gradient under low gravity conditions. During the flight, a sample of a sodium borate melt with a specific bubble array, contained in a platinum/fused silica cell, was subjected to a well defined temperature gradient for more than 4 minutes. Photographs taken at one second intervals during the experiment clearly show that the bubbles move toward the hot spot on the platinum heater strip. This result is consistent with the predictions of the theory of thermocapillary driven bubble motion.

Smith, H. D.↗
DOE OSTI2021

In-situ synchrotron X-ray micro-diffraction investigation of ultra-low-strain deformation microstructure in laminated Ti-Al composites

An ultra-low-strain deformation microstructure was revealed for the first time non-destructively in the bulk interior of an annealed laminated Ti-Al composite-in the "fully recrystallized" Al layer-by a synchrotron-based micro-diffraction technique, namely differential aperture X-ray microscopy (DAXM), through real space mapping with a very high angular resolution (0.01 degrees). This ultra-low-strain deformation microstructure was found to result from the thermal stress, induced during cooling after annealing, due to the different coefficients of thermal expansion for the Ti and Al layers. The annealed sample was further tensile deformed to a strain of 1.66% and followed by in situ DAXM and analyzed by various misorientation mapping methods. The results pointed to the important effects of the initial microstructure and the interface constraint, as well as the grain size and crystal orientation, on the plastic deformation. A gradient in dislocation density from the layer interface to the center was found in the Al layer of the annealed sample, and this gradient increased slightly during tensile deformation. The variation of the dislocation density was further discussed based on the activation and interaction of dislocations in grains of different sizes and orientations during plastic deformation. The findings of this study provided valuable insights in understanding the constraint effect of the laminated metal composite and the design of novel composite materials. (c) 2020 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.

Crystallographic orientation↗
DOE OSTI2025

Efficient Training of Deep Neural Operator Networks via Randomized Sampling

Neural operators (NOs) employ deep neural networks to learn the mappings between infinitedimensional function spaces. Deep operator network (DeepONet), a popular NO architecture, has demonstrated success in the real-time prediction of complex dynamics across various scientific and engineering applications. In this work, we introduce a random sampling technique to be adopted during the training of DeepONet, aimed at improving the generalization ability of the model, while significantly reducing the computational time. The proposed approach targets the trunk network of the DeepONet model that outputs the basis functions corresponding to the spatiotemporal locations of the bounded domain on which the physical system is defined. While constructing the loss function, DeepONet training traditionally considers a uniform grid of spatiotemporal points at which all the output functions are evaluated for each iteration. This approach leads to a larger batch size, resulting in poor generalization and increased memory demands, due to the limitations of the stochastic gradient descent (SGD) optimizer. The proposed random sampling over the inputs of the trunk net mitigates these challenges, improving generalization and reducing the memory requirements during training, resulting in significant computational gains. We validate our hypothesis through three benchmark examples, demonstrating substantial reductions in training time while achieving comparable or lower overall test errors relative to the traditional training approach. Our results indicate that incorporating randomization in the trunk network inputs during training enhances the efficiency and robustness of DeepONet, offering a promising avenue for improving the framework’s performance in modeling complex physical systems.

Karumuri, Sharmila [Department of Civil & Systems ↗
DOE OSTI2021

Laplacian Smoothing Stochastic Gradient Markov Chain Monte Carlo

As an important Markov chain Monte Carlo (MCMC) method, the stochastic gradient Langevin dynamics (SGLD) algorithm has achieved great success in Bayesian learning and posterior sampling. Furthermore, SGLD typically suffers from a slow convergence rate due to its large variance caused by the stochastic gradient. In order to alleviate these drawbacks, we leverage the recently developed Laplacian smoothing technique and propose a Laplacian smoothing stochastic gradient Langevin dynamics (LS-SGLD) algorithm. We prove that for sampling from both log-concave and non-log-concave densities, LS-SGLD achieves strictly smaller discretization error in 2-Wasserstein distance, although its mixing rate can be slightly slower. Experiments on both synthetic and real datasets verify our theoretical results and demonstrate the superior performance of LS-SGLD on different machine learning tasks including posterior sampling, Bayesian logistic regression, and training Bayesian convolutional neural networks.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS2014

Connecting Returned Apollo Soils and Remote Sensing: Application to the Diviner Lunar Radiometer

The Diviner Lunar Radiometer, onboard NASA's Lunar Reconnaissance Orbiter, has produced the first global, high resolution, thermal infrared observations of an airless body. The Moon, which is the most accessible member of this most abundant class of solar system objects, is also the only body for which we have extraterrestrial samples with known spatial context, returned Apollo samples. Here we present the results of a comprehensive study to reproduce an accurate simulated lunar environment, evaluate the most appropriate sample and measurement conditions, collect thermal infrared spectra of a representative suite of Apollo soils, and correlate them with Diviner observations of the lunar surface. It has been established previously that thermal infrared spectra measured in simulated lunar environment (SLE) are significantly altered from spectra measured under terrestrial or martian conditions. The data presented here were collected at the University of Oxford Simulated Lunar Environment Chamber (SLEC). In SLEC, we simulate the lunar environment by: (1) pumping the chamber to vacuum pressures (less than 10‐4 mbar) sufficient to simulate lunar heat transport processes within the sample, (2) cooling the chamber with liquid nitrogen to simulate radiation to the cold space environment, and (3) heating the samples with heaters and lamp to set‐up thermal gradients similar to those experienced in the upper hundreds of microns of the lunar surface. We then conducted a comprehensive suite of experiments using different sample preparation and heating conditions on Apollo soils 15071 (maria) and 67701 (highland) and compared the results to Diviner noontime data to select the optimal experimental conditions. This study includes thermal infrared SLE measurements of 10084 (A11 - LM), 12001 (A12 - LM), 14259 (A14 - LM), 15071 (A15 - S1), 15601 (A15 - S9a), 61141 (A16 - S1), 66031 (A16 - S6), 67701 (A16 - S11), and 70181 (A17 - LM). The Diviner dataset includes all six Apollo sites at approximately 200 m spatial resolution We find that analyses of Diviner observations of individual sampling stations and SLE measurements returned Apollo soils show good agreement, while comparisons to thermal infrared reflectance under ambient conditions do not agree well, which underscores the need for SLE measurements and validates the Diviner compositional measurement technique.

Greenhagen, B. T.↗
NASA NTRS1976

Testing of typical spacecraft materials in a simulated substorm environment

An investigation to determine spacecraft materials characteristics under charging conditions was conducted at the Lewis Research Center substorm simulation facility. The test specimens were spacecraft paints, silvered Teflon, thermal blankets and solar array segments. The samples, ranging in size from 300 to 1,000 sq cm were exposed to monoenergetic electron energies from 2 to 20 keV at a current density of 1 nA sq cm. The samples generally behaved as capacitors with strong voltage gradients at their edges. The charging characteristics of the silvered Teflon, Kapton, and solar cell covers were controlled by the secondary emission characteristics. Insulators that did not discharge were the spacecraft paints and the quartz fiber cloth thermal blanket sample. All other samples did experience discharges when the surface voltage reached -8 to -16 kV. The discharges were photographed. The breakdown voltage for each sample was determined and the average energy lost in the discharge was computed.

Stephens, N. J.↗
NASA NTRS1977

Testing of typical spacecraft materials in a simulated substorm environment

The test specimens were spacecraft paints, silvered Teflon, thermal blankets, and solar array segments. The samples, ranging in size from 300 to 1000 sq cm were exposed to monoenergetic electron energies from 2 to 20 keV at a current density of 1 NA/sq cm. The samples generally behaved as capacitors with strong voltage gradient at their edges. The charging characteristics of the silvered Teflon, Kapton, and solar cell covers were controlled by the secondary emission characteristics. Insulators that did not discharge were the spacecraft paints and the quartz fiber cloth thermal blanket sample. All other samples did experience discharges when the surface voltage reached -8 to -16kV. The discharges were photographed. The breakdown voltage for each sample was determined and the average energy lost in the discharge was computed.

Stevens, N. J.↗
DOE OSTI2021

Machine Learning-Based Classification of Lignocellulosic Biomass from Pyrolysis-Molecular Beam Mass Spectrometry Data

High-throughput analysis of biomass is necessary to ensure consistent and uniform feedstocks for agricultural and bioenergy applications and is needed to inform genomics and systems biology models. Pyrolysis followed by mass spectrometry such as molecular beam mass spectrometry (py-MBMS) analyses are becoming increasingly popular for the rapid analysis of biomass cell wall composition and typically require the use of different data analysis tools depending on the need and application. Here, the authors report the py-MBMS analysis of several types of lignocellulosic biomass to gain an understanding of spectral patterns and variation with associated biomass composition and use machine learning approaches to classify, differentiate, and predict biomass types on the basis of py-MBMS spectra. Py-MBMS spectra were also corrected for instrumental variance using generalized linear modeling (GLM) based on the use of select ions relative abundances as spike-in controls. Machine learning classification algorithms e.g., random forest, k-nearest neighbor, decision tree, Gaussian Naïve Bayes, gradient boosting, and multilayer perceptron classifiers were used. The k-nearest neighbors (k-NN) classifier generally performed the best for classifications using raw spectral data, and the decision tree classifier performed the worst. After normalization of spectra to account for instrumental variance, all the classifiers had comparable and generally acceptable performance for predicting the biomass types, although the k-NN and decision tree classifiers were not as accurate for prediction of specific sample types. Gaussian Naïve Bayes (GNB) and extreme gradient boosting (XGB) classifiers performed better than the k-NN and the decision tree classifiers for the prediction of biomass mixtures. The data analysis workflow reported here could be applied and extended for comparison of biomass samples of varying types, species, phenotypes, and/or genotypes or subjected to different treatments, environments, etc. to further elucidate the sources of spectral variance, patterns, and to infer compositional information based on spectral analysis, particularly for analysis of data without a priori knowledge of the feedstock composition or identity.

59 BASIC BIOLOGICAL SCIENCES↗
NASA NTRS2016

Primary Dendrite Arm Spacing and Trunk Diameter in Al-7-Weight-Percentage Si Alloy Directionally Solidified Aboard the International Space Station

Under a NASA-ESA collaborative research project, three Al-7-weight-percentage Si samples (MICAST-6, MICAST-7 and MICAST 2-12) were directionally solidified aboard the International Space Station to determine the effect of mitigating convection on the primary dendrite array. The samples were approximately 25 centimeters in length with a diameter of 7.8 millimeter-diameter cylinders that were machined from [100] oriented terrestrially grown dendritic Al-7Si samples and inserted into alumina ampoules within the Sample Cartridge Assembly (SCA) inserts of the Low Gradient Furnace (LGF). The feed rods were partially remelted in space and directionally solidified to effect the [100] dendrite-orientation. MICAST-6 was grown at 5 microns per second for 3.75 centimeters and then at 50 microns per second for its remaining 11.2 centimeters of its length. MICAST-7 was grown at 20 microns per second for 8.5 centimeters and then at 10 microns per second for 9 centimeters of its remaining length. MICAST2-12 was grown at 40 microns per second for 11 centimeters. The thermal gradient at the liquidus temperature varied from 22 to 14 degrees Kelvin per centimeter during growth of MICAST-6, from 26 to 24 degrees Kelvin per centimeter for MICAST-7 and from 33 to 31 degrees Kelvin per centimeter for MICAST2-12. Microstructures on the transverse sections along the sample length were analyzed to determine nearest-neighbor spacing of the primary dendrite arms and trunk diameters of the primary dendrite-arrays. This was done along the lengths where steady-state growth prevailed and also during the transients associated with the speed-changes. The observed nearest-neighbor spacings during steady-state growth of the MICAST samples show a very good agreement with predictions from the Hunt-Lu primary spacing model for diffusion controlled growth. The observed primary dendrite trunk diameters during steady-state growth of these samples also agree with predictions from a coarsening-based model. The radial macrosegregation and "steepling" caused by thermosolutal convection during terrestrial growth of the Al-7Si was not observed in the space-grown MICAST samples.

Ghods, M.↗
NASA NTRS1991

Annular denuder systems: Precision and applicability to stratospheric sampling

Annular denuder systems (ADS) are designed to sample both gases and particles from ambient air without cross interference. The ADS's were used to detect chemical gradients over rural forest canopies. Accurate and precise methods for the sampling and analysis of low concentrations of gaseous and particulate species over short sampling periods were developed. Typical analytical precision data for nitrate and sulfate are shown. Both sampling and analytical precision data for a 2 hour test using four collocated ADS's are presented. On this particular day, the overall precision for HNO3 was unusually poor. The overall precision for the major constituents of interest is generally about 3 to 5 percent for typical ambient concentrations. The ADS's could potentially be used to determine the concentrations of HNO3 and fine-particle nitrate in the stratosphere, and to simultaneously characterize the chemical composition of stratospheric aerosols. Assuming 50 mb pressure, a 2 hour sampling time, a 50 Lpm sampling rate, and a low blank value, an overal precision of about 15 percent should be possible for 1 ppbv of HNO3 or nitrate, which would include nitric-acid trihydrate (NAT); and pump-driven systems could accommodate more complex systems involving denuders and filter packs in sequence. Alternate coatings could possibly be developed to selectively sample species such as Cl2 and ClNO3. The ADS's might also be useful in laboratory studies of heterogeneous chemical reactions involving NAT aerosols.

Randtke, Stephen J.↗
DOE OSTI2020

Probabilistic Data-Driven Sampling via Multi-Criteria Importance Analysis

Although supercomputers are becoming increasingly powerful, their components have thus far not scaled proportionately. Compute power is growing enormously and is enabling finely resolved simulations that produce never-before-seen features. However, I/O capabilities lag by orders of magnitude, which means only a fraction of the simulation data can be stored for post hoc analysis. Prespecified plans for saving features and quantities of interest do not work for features that have not been seen before. Data-driven intelligent sampling schemes are needed to detect and save important parts of the simulation while it is running. Here, we propose a novel sampling scheme that reduces the size of the data by orders-of-magnitude while still preserving important regions. The approach we develop selects points with unusual data values and high gradients. Finally, we demonstrate that our approach outperforms traditional sampling schemes on a number of tasks.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2023

Building a genome-based understanding of bacterial pH preferences

The environmental preferences of many microbes remain undetermined. This is the case for bacterial pH preferences, which can be difficult to predict a priori despite the importance of pH as a factor structuring bacterial communities in many systems. We compiled data on bacterial distributions from five datasets spanning pH gradients in soil and freshwater systems (1470 samples), quantified the pH preferences of bacterial taxa across these datasets, and compiled genomic data from representative bacterial taxa. While taxonomic and phylogenetic information were generally poor predictors of bacterial pH preferences, we identified genes consistently associated with pH preference across environments. We then developed and validated a machine learning model to estimate bacterial pH preferences from genomic information alone, a model that could aid in the selection of microbial inoculants, improve species distribution models, or help design effective cultivation strategies. More generally, we demonstrate the value of combining biogeographic and genomic data to infer and predict the environmental preferences of diverse bacterial taxa.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2022

Metatranscriptomic Sequencing of Winter and Spring Planktonic Communities from Lake Erie, a Laurentian Great Lake

Previous reports suggest planktonic and under-ice winter microbial communities in Lake Erie are dominated by diatoms. Here, we report the assembled metatranscriptomes of 79 Lake Erie surface water microbial communities spanning both the winter (28 samples) and spring (51 samples) months over spatial, temporal, and climatic gradients in 2019 through 2020.

Zepernick, Brittany N.↗
DOE OSTI2024

A compact furnace to support in situ neutron imaging of hydrogen dynamics in yttrium hydride moderators

A compact, nuclear microreactor that utilizes low-enriched uranium fuel is a promising solution to meet U.S. energy demands in nonconventional nuclear markets such as remote and decentralized energy grids. Yttrium hydride (YHx) is a potential moderator material for a microreactor design that reduces the amount of required fuel and provides superior retention of hydrogen at high reactor operating temperatures. Hydrogen diffusion properties in YHx are highly sought after for computer model validation and reactor prototyping. To characterize hydrogen diffusion, a compact dual-zone furnace was developed at Los Alamos National Laboratory and analyzed via neutron imaging at the Los Alamos Neutron Science Center (LANSCE). The goal of these measurements is to assess hydrogen diffusion in YHx samples as a function of applied temperature gradients. Included herein is recent progress in technique and furnace developments as well as initial results from concentration- and temperature-gradient measurements at LANSCE.

Torres, James↗
DOE OSTI2022

Modeling O and N Alloying in Nb for SRF Applications

N and O-alloyed superconducting radio frequency cavities exhibit extraordinary quality factors. Developing diffusion models that describe interstitial N and O in Nb is important for optimizing alloyed cavity quality factors and accelerating gradients. N and O-alloyed Nb samples are examined with SEM AND SIMS and their diffusion profiles modeled.

Lechner, Eric↗
NASA NTRS1972

Davidon-Broyden rank-one minimization methods in Hilbert space with application to optimal control problems

The Davidon-Broyden class of rank one, quasi-Newton minimization methods is extended from Euclidean spaces to infinite-dimensional, real Hilbert spaces. For several techniques of choosing the step size, conditions are found which assure convergence of the associated iterates to the location of the minimum of a positive definite quadratic functional. For those techniques, convergence is achieved without the problem of the computation of a one-dimensional minimum at each iteration. The application of this class of minimization methods for the direct computation of the solution of an optimal control problem is outlined. The performance of various members of the class are compared by solving a sample optimal control problem. Finally, the sample problem is solved by other known gradient methods, and the results are compared with those obtained with the rank one quasi-Newton methods.

Straeter, T. A.↗
NASA NTRS1990

Global and radial variations in the efficiency of massive star formation among galaxies

In order to determine the regions within galaxies which give rise to the most efficient star formation and to test the hypothesis that galaxies with high infrared luminosities per unit molecular mass are efficiently producing high mass stars, researchers have undertaken an H alpha imaging survey in galaxies whose CO distributions have been measured as part of the Five College Radio Astronomy Observatory (FCRAO) Extragalactic CO Survey. From these images researchers have derived global H alpha fluxes and distributions for comparison with far infrared radiation (FIR) fluxes and CO fluxes and distributions. Here, researchers present results on the global massive star formation efficiency (SFE = L sub H sub alpha/M(H2)) as a function of morphological type and environment, and on the radial distribution of the SFE within both peculiar and isolated galaxies. On the basis of comparison of the global L sub H sub alpha/M(H2) and L sub FIR/M(H2) for 111 galaxies, researchers conclude that environment rather than morphological type has the strongest effect on the global efficiency of massive star formation. Based on their study of a small sample, they find that the largest radial gradients are observed in the interacting/peculiar galaxies, indicating that environment affects the star formation efficiency within galaxies as well.

Allen, Lori E.↗