Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “factor analysis”

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 991 records · Page 55

Physics and chemistry from parsimonious representations: image analysis via invariant variational autoencoders

Electron, optical, and scanning probe microscopy methods are generating ever increasing volume of image data containing information on atomic and mesoscale structures and functionalities. This necessitates the development of the machine learning methods for discovery of physical and chemical phenomena from the data, such as manifestations of symmetry breaking phenomena in electron and scanning tunneling microscopy images, or variability of the nanoparticles. Variational autoencoders (VAEs) are emerging as a powerful paradigm for the unsupervised data analysis, allowing to disentangle the factors of variability and discover optimal parsimonious representation. Here, we summarize recent developments in VAEs, covering the basic principles and intuition behind the VAEs. The invariant VAEs are introduced as an approach to accommodate scale and translation invariances present in imaging data and separate known factors of variations from the ones to be discovered. We further describe the opportunities enabled by the control over VAE architecture, including conditional, semi-supervised, and joint VAEs. Several case studies of VAE applications for toy models and experimental datasets in Scanning Transmission Electron Microscopy are discussed, emphasizing the deep connection between VAE and basic physical principles. Python codes and datasets discussed in this article are available at https://github.com/saimani5/VAE-tutorials and can be used by researchers as an application guide when applying these to their own datasets.

36 MATERIALS SCIENCE↗

Stereo-selective partitioning of translation-to-internal energy conversion in gas ensembles

A recent computational study of translation-to-internal energy transfer to H{sub 2} (v = 0,j = 0), hereinafter denoted H{sub 2} (0;0), in a bath of H atoms [A. J. McCaffery and R. J. Marsh, J. Chem. Phys. 139, 234310 (2013)] revealed an unexpected energy partitioning in which the H{sub 2} vibrational temperature greatly exceeds that of rotation. This occurs despite rotation and vibration distributions being close to Boltzmann from early in ensemble evolution. In this work, the study is extended to include H{sub 2} (0;0), O{sub 2} (0;0), and HF (0;0) in a wide range of atomic bath gases comprising some 22 ensembles in all. Translation-to-internal energy conversion in the systems studied was found to be relatively inefficient, falling approximately with (√μ′){sup −1} as bath gas mass increases, where μ′ is the reduced mass of the diatomic–bath gas pair. In all 22 systems studied, T{sub v} exceeds T{sub r} – by a factor > 4 for some pairs. Analysis of the constraints that influence (0;0) → (1;j) excitation for each diatomic–atom pair in momentum–angular momentum space demonstrates that a vibrational preference results from energy constraints that limit permitted collision trajectories to those of low effective impact parameter, i.e., to those that are axial or near axial on impact with the Newton surface. This implies that a steric constraint is an inherent feature of vibration-rotation excitation and arises because momentum and energy barriers must be overcome before rotational states may be populated in the higher vibrational level.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

IMAGING SPECTROPOLARIMETRY WITH IBIS. II. ON THE FINE STRUCTURE OF G-BAND BRIGHT FEATURES

We present new results from first observations of the quiet solar photosphere performed through the Interferometric BIdimensional Spectrometer (IBIS) in spectropolarimetric mode. IBIS allowed us to measure the four Stokes parameters in the Fe I 630.15 nm and Fe I 630.25 nm lines with high spatial and spectral resolutions for 53 minutes; the polarimetric sensitivity achieved by the instrument is 3 x 10{sup -3} the continuum intensity level. We focus on the correlation which emerges between the G-band bright feature brightness and magnetic filling factor of {approx}10{sup 3} G (kG) fields derived by inverting the Stokes I and V profiles. In more detail, we present the correlation first in a pixel-by-pixel study of a {approx_equal}3'' wide bright feature (a small network patch) and then we show that such a result can be extended to all the bright features found in the data set at any instant of the time sequence. The higher the kilo-Gauss filling factor associated to a feature, the higher the brightness of the feature itself. Filling factors up to {approx_equal}35% are obtained for the brightest features. Considering the values of the filling factors derived from the inversion analysis of spectropolarimetric data and the brightness variation observed in the G-band data we put forward an upper limit for the smallest scale over which magnetic flux concentrations in intergranular lanes produce a G-band brightness enhancement ({approx_equal}0.''1). Moreover, the brightness saturation observed for feature sizes comparable to the resolution of the observations is compatible with the large G-band bright features being clusters of sub-arcsecond bright points. This conclusion deserves to be confirmed by forthcoming spectropolarimetric observations at higher spatial resolution.

79 ASTRONOMY AND ASTROPHYSICS↗

TRANSIENT SUPERSTRONG CORONAL LINES AND BROAD BUMPS IN THE GALAXY SDSS J074820.67+471214.3

Variable superstrong coronal emission lines were observed in the spectrum of one galaxy, SDSS J095209.56+214313.3, and their enigmatic origin remains controversial. In this paper, we report the detection of variable broad emission bumps reminiscent of a supernova (SN) II-Plateau spectra taken a few days after the shock breakout in a second galaxy with variable superstrong coronal lines, SDSS J074820.67+471214.3. The coronal line spectrum shows unprecedentedly high ionization with superstrong [Fe X]{lambda}6376, [Fe XI]{lambda}7894, [Fe XIV]{lambda}5304, [S XII]{lambda}7612, and [Ar XIV]{lambda}4414, but without detectable optical [Fe VII] line emission. The coronal line luminosities are similar to those observed in bright Seyfert galaxies and 20 times more luminous than those reported in the hottest Type IIn SN 2005ip inferred from its strong coronal lines. The coronal lines ({sigma} {approx} 120-240 km s{sup -1}) are much broader than the narrow emission lines ({sigma} {approx} 40 km s{sup -1}) from the star-forming regions in the galaxy, but are nearly at the same systematic redshift. We also detected a variable non-stellar continuum emission from its Sloan Digital Sky Survey spectroscopy and Galaxy Evolution Explorer photometry. In the follow-up spectra taken 4-5 years later, the coronal lines, SN-like feature, and non-stellar continuum disappeared, while the [O III]{lambda}5007 intensity increased by a factor of about 10. Our analysis suggests that the coronal line region should be at least 10 light days in size and should be powered either by a steady ionizing source with a soft X-ray luminosity of at least a few 10{sup 42} erg s{sup -1} or by a very luminous soft X-ray outburst. These findings can be more naturally explained by a star tidally disrupted by the central black hole than by an SN explosion. The similarity of the coronal line variability trend observed in the two galaxies suggests that the two transient events have the same origin, with SDSS J074820.67+471214.3 being caught at an earlier stage by the spectroscopic observation.

79 ASTRONOMY AND ASTROPHYSICS↗

Star formation and black hole growth at z ≅ 4.8

We report Herschel/SPIRE, Spitzer and Wide-field Infrared Survey Explorer observations of 44 z ≅ 4.8 optically selected active galactic nuclei (AGNs). This flux-limited sample contains the highest mass black holes (BHs) at this redshift. Ten of the objects were detected by Herschel and five show emission that is not clearly associated with the AGNs. The star formation (SF) luminosity (L{sub SF}) obtained by fitting the spectral energy distribution (SED) with standard SF templates, taking into account AGN contribution, is in the range 10{sup 46.62}-10{sup 47.21} erg s{sup –1} corresponding to SF rates of 1090-4240 M{sub ☉} yr{sup –1}. Fitting with very luminous submillimeter galaxy SEDs gives SF rates that are smaller by 0.05 dex when using all bands and 0.1 dex when ignoring the 250 μm band. A 40 K graybody fits to only the 500 μm fluxes reduce L{sub SF} by about a factor of two. A stacking analysis of 29 undetected sources gives significant signals in all three bands. A SF template fit indicates L{sub SF} = 10{sup 46.19-46.23} erg s{sup –1} depending on the assumed AGN contribution. A 40 K fit to the stacked 500 μm flux gives L{sub SF} = 10{sup 45.95} erg s{sup –1}. The mean BH mass (M{sub BH}) and AGN luminosity (L{sub AGN}) of the detected sources are significantly higher than those of the undetected ones. The spectral differences are seen all the way from UV to far infrared wavelengths. The mean optical-UV spectra are similar to those predicted for thin accretion disks around BHs with similar masses and accretion rates. We suggest two alternative explanations to the correlation of L{sub SF}, L{sub AGN} and M{sub BH}, one involving no AGN feedback and the second involving moderate feedback that affects, but does not totally quench, SF in three-quarters of the sources. We compare our L{sub SF} and L{sub AGN} to lower redshift samples and show a new correlation between L{sub SF} and M{sub BH}. We also examine several rather speculative ideas about the host galaxy properties including the possibility that the detected sources are above the SF mass sequence (MS) at z ≅ 4.8, perhaps in mergers, and most of the undetected sources are on the MS.

79 ASTRONOMY AND ASTROPHYSICS↗

THE TURBULENCE VELOCITY POWER SPECTRUM OF NEUTRAL HYDROGEN IN THE SMALL MAGELLANIC CLOUD

We present the results of the Velocity Coordinate Spectrum (VCS) technique to calculate the velocity power spectrum of turbulence in the Small Magellanic Cloud (SMC) in 21 cm emission. We present an updated version of the VCS technique that takes into account regular motions, which is an important factor in our SMC VCS analysis. We have obtained a velocity spectral index of −3.85, a cold phase sonic Mach number of 5.6, and an injection scale of 2.3 kpc. The spectral index is steeper than the Kolmogorov index, which is expected for shock-dominated turbulence. The injection scale of 2.3 kpc suggests that HI supershells or tidal interactions with the Large Magellanic Cloud are the dominant drivers of turbulence in this dwarf galaxy. This implies that turbulence may be driven by multiple mechanisms in galaxies and that galaxy–galaxy interactions may play an important role in addition to supernova feedback.

79 ASTRONOMY AND ASTROPHYSICS↗

Effects of primordial black holes on IGM history

Currently the asteroid mass window (mass ∼ 10$^{17}$- 10$^{21}$ grams) remains unconstrained for Primordial Black Holes (PBHs) to make up all of the dark matter content of the universe. Given these PBHs have very small masses, their Hawking temperature can be up to hundreds of keV. This study investigates the potential impacts of PBH Hawking radiation on the intergalactic medium from z ∼ 800-25, namely studying the ionization history, kinetic gas temperature, and ultimately the 21 cm signature. We find that for masses on the low edge of the asteroid mass window, there are up two orders of magnitude increases in the ionization fraction and kinetic gas temperature by redshift 25, and the 21 cm spin temperature can differ from non-PBH cosmology by factors of a few. This analysis results in maximum differential brightness temperatures of +17 mK for our lightest PBH masses of 2.12 × 10$^{16}$g. We also show maximal 53 mK discrepancies in differential brightness temperatures between our PBH and non-PBH cosmologies for our lightest PBH mass, while our heaviest PBH mass of 1.65 × 10$^{17}$g shows only 0.5 mK variations. We find the Hawking-radiated electrons and positrons are instrumental in driving these IGM modifications. This study shows the necessity for a rigorous treatment of Hawking radiation in PBH cosmological observables from the dark ages through cosmic dawn.

dark matter theory↗

Performance of GeantV EM Physics Models

The recent progress in parallel hardware architectures with deeper vector pipelines or many-cores technologies brings opportunities for HEP experiments to take advantage of SIMD and SIMT computing models. Launched in 2013, the GeantV project studies performance gains in propagating multiple particles in parallel, improving instruction throughput and data locality in HEP event simulation on modern parallel hardware architecture. Due to the complexity of geometry description and physics algorithms of a typical HEP application, performance analysis is indispensable in identifying factors limiting parallel execution. In this report, we will present design considerations and preliminary computing performance of GeantV physics models on coprocessors (Intel Xeon Phi and NVidia GPUs) as well as on mainstream CPUs.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

RELATING CHANGES IN COMETARY ROTATION TO ACTIVITY: CURRENT STATUS AND APPLICATIONS TO COMET C/2012 S1 (ISON)

We introduce a parameter, X, to predict the changes in the rotational period of a comet in terms of the rotational period itself, the nuclear radius, and the orbital characteristics. We show that X should be a constant if the bulk densities and shapes of nuclei are nearly identical and the activity patterns are similar for all comets. For four nuclei for which rotational changes are well documented, despite the nearly factor 30 variation observed among the effective active fractions of these comets, X is constant to within a factor two. We present an analysis for the sungrazing comet C/2012 S1 (ISON) to explore what rotational changes it could undergo during the upcoming perihelion passage where its perihelion distance will be ∼2.7 solar radii. When close to the Sun, barring a catastrophic disruption of the nucleus, the activity of ISON will be sufficiently strong to put the nucleus into a non-principal-axis rotational state and observable changes to the rotational period should also occur. Additional causes for rotational state changes near perihelion for ISON are tidal torques caused by the Sun and the significant mass loss due to a number of mechanisms resulting in alterations to the moments of inertia of the nucleus.

79 ASTRONOMY AND ASTROPHYSICS↗

The cross-section measurement for the 3 H($e, e'K^+)nnΛ$ reaction

The small binding energy of the hypertriton leads to predictions of the non-existence of bound hypernuclei for isotriplet three-body systems such as nnΛ. However, invariant mass spectroscopy at GSI has reported events that may be interpreted as the bound nnΛ state. The nnΛ state was sought by missing-mass spectroscopy via the (e, e'K + ) reaction at Jefferson Lab’s experimental Hall A. The present experiment has higher sensitivity to the nnΛ-state investigation in terms of better precision by a factor of about three. The analysis shown in this article focuses on the derivation of the reaction cross-section for the 3 H(γ*, K + )X reaction. Events that were detected in an acceptance, where a Monte Carlo simulation could reproduce the data well (|δp/p| < 4%), were analyzed to minimize the systematic uncertainty. No significant structures were observed with the acceptance cuts, and the upper limits of the production cross-section of the nnΛ state were obtained to be 21 and $31 nbsr -1 at the 90% confidence level when theoretical predictions of (-B Λ , Γ) = (0.25, 0.8) MeV and (0.55, 4.7) MeV, respectively, were assumed. The cross-section result provides valuable information for examining the existence of nnΛ.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Correlation between temperature variations of static and dynamic properties in glass-forming liquids

In this detailed analysis of the static structure factor S(Q) in several glass-forming liquids we show that the temperature variations of the width of the main diffraction peak Q(T ) correlate with the fragility of these liquids. Our observation suggests a direct connection between rather subtle structural changes and sharp slowing down of structural relaxation in glass-forming liquids. We also show that this observation can be rationalized using the Adam-Gibbs approach, through a connection between temperature variations of structural correlation length, lc 2 /Q, and the size of cooperatively rearranging regions.

36 MATERIALS SCIENCE↗

Isomeric states in the neutron-rich 𝑁 = 108 nuclei 178 Yb and 179 Lu

A multinucleon transfer reaction with a 136 Xe beam and a gold-backed 176 Yb target populated isomers in N = 108 178 Yb and 179 Lu. The single decay from the t 1/2 = 203(16) ns isomer in 178 Yb to the I π = 4 + level in the ground-state band suggests a I π = 4 − assignment, and multiquasiparticle calculations associate this state with the K π = 4 − , ν 2 (1/2[510], 9/2[624]) configuration. At least two isomers were observed in 179 Lu. One likely couples the K π = 4 − configuration found in 178 Yb to the 9/2[514] quasiproton orbital to form a K π = 17/2 + level (t 1/2 = 1.0(4) µs) which primarily decays into the rotational band based on the 9/2[514] state. A second isomer was observed to decay into this same band, but at higher energy and spin. Based on its feeding pattern, its similarity to an isomer found in 181 Ta, and a multiquasiparticle calculation, it is proposed to correspond to a K π = 29/2 − state based on the π9/2[514]ν 2 (9/2[624], 11/2[615]) configuration. As a result, an analysis of the reduced hindrance factors for negative-parity, two-quasineutron isomers in the rare-earth region is presented as well.

150 ≤ A ≤ 189↗

Toward a Holistic Performance Evaluation of Large Language Models Across Diverse AI Accelerators

Artificial intelligence (AI) methods have become critical in scientific applications to help accelerate scientific discovery. Large language models (LLMs) are being considered a promising approach to address some challenging problems because of their superior generalization capabilities across domains. The effectiveness of the models and the accuracy of the applications are contingent upon their efficient execution on the underlying hardware infrastructure. Specialized Al accelerator hardware systems have recently become available for accelerating Al applications. However, the comparative performance of these AI accelerators on large language models has not been previously studied. In this paper, we systematically study LLMs on multiple AI accelerators and GPUs and evaluate their performance characteristics for these models. We evaluate these systems with (i) a micro-benchmark using a core transformer block, (ii) a GPT-2 model, and (iii) an 1,I,M-driven science use case, GenSLM. We present our findings and analyses of the models' performance to better understand the intrinsic capabilities of AI accelerators. Furthermore, our analysis takes into account key factors such as sequence lengths, scaling behavior, and sensitivity to gradient accumulation steps.

Emani, Murali↗

Cy-Phy ADS: Cyber Physical Anomaly Detection Framework for EV Charging Systems

Today’s large-scale Electric Vehicle (EV) infrastructures are heavily dependent on information communication technologies to maintain their operation and to support communication within sub-system components as well as the outside world. These technologies are vulnerable to various cyber and physical threats. Timely identification and mitigation of these threats are critical for improving human safety, avoiding economic losses, and preventing catastrophic system failures. By addressing this, our work presents a ResNet Autoencoder (AE) based Cyber-Physical Anomaly Detection System (Cy-Phy ADS) for detecting anomalies in EV Controller Area Network (CAN) protocol communication. It consists of four main components: Cyber-Physical Feature Extractor, ResNet AE-based Anomaly Detection Framework, Cyber-Physical Health Metric (CPHM), and Visualization Dashboard. The presented framework was trained and tested using CAN data collected from the EV charging system testbed at the Idaho National Laboratory. The presented Cy-Phy ADS compared against six widely used unsupervised anomaly detection algorithms: One Class Support Vector Machine (OCSVM), Variational Autoencoder (VAE), LSTM Autoencoder (LSTM AE), Isolation Forest (IForest), Principle Component Analysis (PCA) and Local Outlier Factor (LOF). Here the presented approach showed the highest accuracy among the compared methods. Further, the proposed approach showed comparable performance in terms of precision, F1, and False positive rate. It also showed the lowest training and inference time compared to the neural network-based baseline algorithms compared against with. Additionally, the Cy-Phy ADS has advantages such as unsupervised training, the ability to provide a holistic metric for system health characterization, and non-linear feature extraction.

99 GENERAL AND MISCELLANEOUS↗

ShiftNMFk 1.2

The ShiftNMFk1.2 code, or as we call it, GreenNMFk, represents a hybrid algorithm combining unsupervised adaptive machine learning and Green's function inverse method. GreenNMFk allows an efficient and high performance de-mixing and feature extraction of a multitude of nonnegative signals that change their shape propagating through the medium. The signals are mixed and recorded by a network of uncorrelated sensors. The code couples Non-negative Matrix Factorization (NMF) and inverse-analysis Green's functions method. GreenNMF synergistically performs decomposition of the recorded mixtures, finds the number of the unknown sources and uses the Green's function of the governing partial differential equation to identify the unknown sources and their charecteristics. GreenNMF can be applied directly to any problem controlled by a known partial-differential parabolic equation where mixtures of an unknown number of sources are measured at multiple locations. Full GreenNMFk method is a subject LANL U.S. Patent application S133364.000 August, 2017. The ShiftNMFk 1.2 version here is a toy version of this method that can work with a limited number of unknown sources (4 or less).

Alexandrov, BoianS.↗

Quality scores for 32,000 genomes

More than 80% of the microbial genomes in GenBank are of ‘draft’ quality (12,553 draft vs. 2,679 finished, as of October, 2013). In this study, we have examined all the microbial DNA sequences available for complete, draft, and Sequence Read Archive genomes in GenBank as well as three other major public databases, and assigned quality scores for more than 30,000 prokaryotic genome sequences. Scores were assigned using four categories: the completeness of the assembly, the presence of full-length rRNA genes, tRNA composition and the presence of a set of 102 conserved genes in prokaryotes. Most (~88%) of the genomes had quality scores of 0.8 or better and can be safely used for standard comparative genomics analysis. We compared genomes across factors that may influence the score. We found that although sequencing depth coverage of over 100x did not ensure a better score, sequencing read length was a better indicator of sequencing quality. With few exceptions, most of the 30,000 genomes have nearly all the 102 essential genes. The score can be used to set thresholds for screening data when analyzing “all published genomes” and reference data is either not available or not applicable. The scores highlighted organisms for which commonly used tools do not perform well. This information can be used to improve tools and to serve a broad group of users as more diverse organisms are sequenced. Finally and unexpectedly, the comparison of predicted tRNAs across 15,000 high quality genomes showed that anticodons beginning with an ‘A’ (codons ending with a ‘U’) are almost non-existent, with the exception of one arginine codon (CGU); this has been noted previously in the literature for a few genomes, but not with the depth found here.

59 BASIC BIOLOGICAL SCIENCES↗

Preliminary Study of Iodine Gas Removal in Sodium Pools

Potential iodine gas release from failed fuel pins is a critical factor in the source term analysis of oxide fuel-loaded sodium fast reactors (SFRs). The accumulated iodine-containing gas mixtures inside pin plenums are expected to be ejected during pin failures and rise through sodium pool, with potential release of gaseous iodine to the cover gas region. Due to its potential radiological impacts, a proper assessment of iodine behavior is necessary for an accurate source term assessment. Throughout the bubble rise trajectory in the sodium pool, iodine gas is continuously removed or transformed at the bubble interface by diffusion, as the combining reaction between the iodine and sodium to form sodium iodide (NaI) is a chemically preferred process. As the final amount of iodine released from the facility is strongly influenced by the removal phenomenon inside the sodium pool, experiments were previously performed by PNC (Power Reactor and Nuclear Fuel Development Corporation) to provide insight into this phenomenon. To assess the accuracy of present methods for predicting iodine gas removal within sodium pools, several candidate approaches, available in source term analysis codes, have been summarized and evaluated in this study. Spherical cap bubbles and spherical bubbles are considered in accordance with the methods adopted in each approach, and different forms of correlations for major parameters have been implemented in accordance with the original adoptions. Based on the summarized results, important aspects to be considered have been derived.

Decontamination↗