Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Topic modeling”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 1,063 records · Page 59
All sourcesNASA NTRSDOE OSTI
DOE OSTI2019

Lattice Softening Significantly Reduces Thermal Conductivity and Leads to High Thermoelectric Efficiency

The influence of micro/nanostructure on thermal conductivity is a topic of great scientific interest, particularly to thermoelectrics. The current understanding is that structural defects decrease thermal conductivity through phonon scattering where the phonon dispersion and speed of sound are assumed to remain constant. Experimental work on a PbTe model system is presented, which show that the speed of sound linearly decreases with increased internal strain. This softening of the materials lattice completely accounts for the reduction in lattice thermal conductivity, without the introduction of additional phonon scattering mechanisms. Additionally, it is shown that a major contribution to the improvement in the thermoelectric figure of merit (zT > 2) of high-efficiency Na-doped PbTe can be attributed to lattice softening. While inhomogeneous internal strain fields are known to introduce phonon scattering centers, this study demonstrates that internal strain can modify phonon propagation speed as well. This introduces new avenues to control lattice thermal conductivity, beyond phonon scattering. In practice, many engineering materials will exhibit both softening and scattering effects, as is shown in silicon. Here, we shine new light on studies of thermal conductivity in fields of energy materials, microelectronics, and nanoscale heat transfer.

36 MATERIALS SCIENCE↗
DOE OSTI2023

CHQ- SocioEmo: Identifying Social and Emotional Support Needs in Consumer-Health Questions

General public, often called consumers, are increasingly seeking health information online. To be satisfactory, answers to health-related questions often have to go beyond informational needs. Automated approaches to consumer health question answering should be able to recognize the need for social and emotional support. Recently, large scale datasets have addressed the issue of medical question answering and highlighted the challenges associated with question classification from the standpoint of informational needs. However, there is a lack of annotated datasets for the non-informational needs. We introduce a new dataset for non-informational support needs, called CHQ-SocioEmo. The Dataset of Consumer Health Questions was collected from a community question answering forum and annotated with basic emotions and social support needs. This is the first publicly available resource for understanding non-informational support needs in consumer health-related questions online. We benchmark the corpus against multiple state-of-the-art classification models to demonstrate the dataset’s effectiveness.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2017

Intrabasin Variability of East Pacific Tropical Cyclones During ENSO Regulated by Central American Gap Winds

Hurricane Patricia in 2015 was the strongest Pacific hurricane to make landfall in Mexico. Although Patricia fortuitously spared major cities, it reminded us of the threat tropical cyclones (TCs) pose in the eastern North Pacific (ENP) and the importance of improving our understanding and prediction of ENP TCs. Patricia’s intensity and the active 2015 ENP hurricane season have been partially attributed to the strong El Niño in 2015, however there is still a lack of fundamental understanding of the relationship between El Niño-Southern Oscillation (ENSO) and ENP TCs. Here, we demonstrate that ENSO drives intrabasin variability of ENP TCs, with enhanced (reduced) TC frequency in the western portion of the ENP during El Niño (La Niña), but reduced (enhanced) TC frequency in the eastern nearshore area, where landfalling TCs preferentially form. This intrabasin difference is primarily driven by the Central American Gap Winds (CAGW), which intensify (weaken) during El Niño (La Niña), producing low-level anticyclonic (cyclonic) relative vorticity anomalies and thus an unfavorable (favorable) environment for TC genesis. These findings shed new light on the dynamics linking ENP TC activity to ENSO, and highlight the importance of improving CAGW representation in models to make skillful seasonal forecasts of ENP TCs.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2023

Improving deep learning model performance under parametric constraints for materials informatics applications

Abstract Modern machine learning (ML) and deep learning (DL) techniques using high-dimensional data representations have helped accelerate the materials discovery process by efficiently detecting hidden patterns in existing datasets and linking input representations to output properties for a better understanding of the scientific phenomenon. While a deep neural network comprised of fully connected layers has been widely used for materials property prediction, simply creating a deeper model with a large number of layers often faces with vanishing gradient problem, causing a degradation in the performance, thereby limiting usage. In this paper, we study and propose architectural principles to address the question of improving the performance of model training and inference under fixed parametric constraints. Here, we present a general deep-learning framework based on branched residual learning (BRNet) with fully connected layers that can work with any numerical vector-based representation as input to build accurate models to predict materials properties. We perform model training for materials properties using numerical vectors representing different composition-based attributes of the respective materials and compare the performance of the proposed models against traditional ML and existing DL architectures. We find that the proposed models are significantly more accurate than the ML/DL models for all data sizes by using different composition-based attributes as input. Further, branched learning requires fewer parameters and results in faster model training due to better convergence during the training phase than existing neural networks, thereby efficiently building accurate models for predicting materials properties.

36 MATERIALS SCIENCE↗
DOE OSTI2022

A perspective on iron (Fe) in the atmosphere: air quality, climate, and the ocean

As scientists engage in research motivated by climate change and the impacts of pollution on air, water, and human health, we increasingly recognize the need for the scientific community to improve communication and knowledge exchange across disciplines to address pressing and outstanding research questions holistically. Our professional paths have crossed because our research activities focus on the chemical reactivity of Fe-containing minerals in air and water, and at the air-sea interface. (Photo)chemical reactions driven by Fe can take place at the surface of the particles/droplets or within the condensed phase. The extent and rates of these reactions are influenced by water content and biogeochemical activity ubiquitous in these systems. One of these reactions is the production of reactive oxygen species (ROS) that cause damage to respiratory organs. Another is that the reactivity of Fe and organics in aerosol particles alter surficial physicochemical properties that impact aerosol-radiation and aerosol–cloud interactions. Also, upon deposition, aerosol particles influence ocean biogeochemical processes because micronutrients such as Fe or toxic elements such as copper become bioavailable. We provide a perspective on these topics and future research directions on the reactivity of Fe in atmospheric aerosol systems, from sources to short- and long-term impacts at the sinks with emphasis on needs to enhance the predictive power of atmospheric and ocean models.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2025

The glass-ceiling convective regime and the origin and diversity of coronae on Venus

Venus and Earth are rocky planets of roughly the same size and bulk density, yet their surface volcanic and tectonic features appear substantially different. On Venus, the coexistence of large volcanic highlands—interpreted as the surface expression of long-lived mantle plumes—alongside coronae, smaller features thought to be caused by transient thermal diapirs, remains enigmatic. Using two-dimensional numerical models of mantle convection with sharp and broad mineral phase transitions for pyrolite, we show that both scales of upwellings can be generated in a stagnant lid planet with an interior temperature 250 to 400 K warmer than Earth’s. The smaller plumes originate from a ~600 km deep internal layer that exists as a consequence of the different sequence of mineral phase transitions that occur in warmer mantles less processed and differentiated by partial melting and volcanism. Future models that include melting will provide further tests of our hypothesis.

Science & Technology - Other Topics↗
DOE OSTI2011

POSSIBLE EVIDENCE FOR A FISK-TYPE HELIOSPHERIC MAGNETIC FIELD. I. ANALYZING ULYSSES/KET ELECTRON OBSERVATIONS

The propagation of energetic charged particles in the heliospheric magnetic field is one of the fundamental problems in heliophysics. In particular, the structure of the heliospheric magnetic field remains an unsolved problem and is discussed as a controversial topic. The first successful analytic approach to the structure of the heliospheric magnetic field was the Parker field. However, the measurements of the Ulysses spacecraft at high latitudes revealed the possible need for refinements of the existing magnetic field model during solar minimum. Among other reasons, this led to the development of the Fisk field. This approach is highly debated and could not be ruled out with magnetic field measurements so far. A promising method to trace this magnetic field structure is to model the propagation of electrons in the energy range of a few MeV. Employing three-dimensional and time-dependent simulations of the propagation of energetic electrons, this work shows that the influence of a Fisk-type field on the particle transport in the heliosphere leads to characteristic variations of the electron intensities on the timescale of a solar rotation. For the first time it is shown that the Ulysses count rates of 2.5-7 MeV electrons contain the imprint of a Fisk-type heliospheric magnetic field structure. From a comparison of simulation results and the Ulysses count rates, realistic parameters for the Fisk theory are derived. Furthermore, these parameters are used to investigate the modeled relative amplitudes of protons and electrons, including the effects of drifts.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2020

Using Machine Learning Tools to Predict Compressor Stall

Clean energy has become an increasingly important consideration in today’s power systems. As the push for clean energy continues, many coal-fired power plants are being decommissioned in favor of renewable power sources such as wind and solar. However, the intermittent nature of renewables means that dynamic load following traditional power systems is crucial to grid stability. With high flexibility and fast response at a wide range of operating conditions, gas turbine systems are poised to become the main load following component in the power grid. Yet, rapid changes in load can lead to fluid flow instabilities in gas turbine power systems. These instabilities often lead to compressor surge and stall, which are some of the most critical problems facing the safe and efficient operation of compressors in turbomachinery today. Although the topic of compressor surge and stall has been extensively researched, no methods for early prediction have been proven effective. This study explores the utilization of machine learning tools to predict compressor stall. The long short-term memory (LSTM) model, a form of recurrent neural network (RNN), was trained using real compressor stall datasets from a 100 kW recuperated gas turbine power system designed for hybrid configuration. Two variations of the LSTM model, classification and regression, were tested to determine optimal performance. The regression scheme was determined to be the most accurate approach, and a tool for predicting compressor stall was developed using this configuration. Overall, results show that the tool is capable of predicting stalls 5–20 ms before they occur. With a high-speed controller capable of 5 ms time-steps, mitigating action could be taken to prevent compressor stall before it occurs.

42 ENGINEERING↗
DOE OSTI2022

Provably efficient machine learning for quantum many-body problems

Classical machine learning (ML) provides a potentially powerful approach to solving challenging quantum many-body problems in physics and chemistry. However, the advantages of ML over traditional methods have not been firmly established. In this work, we prove that classical ML algorithms can efficiently predict ground-state properties of gapped Hamiltonians after learning from other Hamiltonians in the same quantum phase of matter. By contrast, under a widely accepted conjecture, classical algorithms that do not learn from data cannot achieve the same guarantee. We also prove that classical ML algorithms can efficiently classify a wide range of quantum phases. Extensive numerical experiments corroborate our theoretical results in a variety of scenarios, including Rydberg atom systems, two-dimensional random Heisenberg models, symmetry-protected topological phases, and topologically ordered phases.

Science & Technology - Other Topics↗
DOE OSTI2018

XJB-5-131-mediated improvement in physiology and behaviour of the R6/2 mouse model of Huntington's disease is age- and sex- dependent

We have reported that the radical scavenger XJB-5-131 attenuates or reverses progression of the disease phenotype in the HdhQ(150/150) mouse, a slow onset model of HD. Here, we tested whether XJB-5-131 has beneficial effects in R6/2 mice, a severe early onset model of HD. We found that XJB-5-131 has beneficial effects in R6/2 mice, by delaying features of the motor and histological phenotype. The impact was sex-dependent, with a stronger effect in male mice. XJB-5-131 treatment improved some locomotor deficits in female R6/2 mice, but the effects were, in general, greater in male mice. Chronic treatment of male R6/2 mice with XJB-5-1-131 reduced weight loss, and improved the motor and temperature regulation deficits, especially in male mice. Treatment with XJB-5-131 had no effect on the lifespan of R6/2 mice. Nevertheless, it significantly slowed somatic expansion at 90 days, and reduced the density of inclusions. Our data show that while treatment with XJB-5-131 had complex effects on the phenotype of R6/2 mice, it produced a number of significant improvements in this severe model of HD.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2013

Flavor Dependence of the EMC Effect via Inclusive Electron Scattering from $^{40}$Ca and $^{48}$Ca

The modification of the quark distribution functions in nuclei, the EMC effect, has been a topic of intense experimental and theoretical study since its original observation in 1983. At present there is no uniformly accepted explanation for the origin of the EMC effect and new observables are required to discriminate among the plethora of models put forward. Recent calculations suggest that the up quark distribution in a nucleus is modified differently than the down quark distribution for the case N dne Z. A number of techniques have been proposed to measure this flavor dependence of the EMC effect, including semi-inclusive deep inelastic scattering (SIDIS) and parity violating deep inelastic scattering (PVDIS). In this talk, I will examine the possibility of determining the flavor dependence of the EMC effect via inclusive electron scattering from $^{40}$Ca and $^{48}$Ca.

Gaskell, David↗
DOE OSTI2017

An Overview of Literature Topics Related to Current Concepts, Methods, Tools, and Applications for Cumulative Risk Assessment (2007–2016)

Cumulative risk assessments (CRAs) address combined risks from exposures to multiple chemical and nonchemical stressors and may focus on vulnerable communities or populations. Significant contributions have been made to the development of concepts, methods, and applications for CRA over the past decade. Work in both human health and ecological cumulative risk has advanced in two different contexts. The first context is the effects of chemical mixtures that share common modes of action, or that cause common adverse outcomes. In this context two primary models are used for predicting mixture effects, dose addition or response addition. The second context is evaluating the combined effects of chemical and nonchemical (e.g., radiation, biological, nutritional, economic, psychological, habitat alteration, land-use change, global climate change, and natural disasters) stressors. CRA can be adapted to address risk in many contexts, and this adaptability is reflected in the range in disciplinary perspectives in the published literature. This article presents the results of a literature search and discusses a range of selected work with the intention to give a broad overview of relevant topics and provide a starting point for researchers interested in CRA applications.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2015

MUTZ-3 derived Langerhans cells in human skin equivalents show differential migration and phenotypic plasticity after allergen or irritant exposure

After allergen or irritant exposure, Langerhans cells (LC) undergo phenotypic changes and exit the epidermis. In this study we describe the unique ability of MUTZ-3 derived Langerhans cells (MUTZ-LC) to display similar phenotypic plasticity as their primary counterparts when incorporated into a physiologically relevant full-thickness skin equivalent model (SE-LC). We describe differences and similarities in the mechanisms regulating LC migration and plasticity upon allergen or irritant exposure. The skin equivalent consisted of a reconstructed epidermis containing primary differentiated keratinocytes and CD1a{sup +} MUTZ-LC on a primary fibroblast-populated dermis. Skin equivalents were exposed to a panel of allergens and irritants. Topical exposure to sub-toxic concentrations of allergens (nickel sulfate, resorcinol, cinnamaldehyde) and irritants (Triton X-100, SDS, Tween 80) resulted in LC migration out of the epidermis and into the dermis. Neutralizing antibody to CXCL12 blocked allergen-induced migration, whereas anti-CCL5 blocked irritant-induced migration. In contrast to allergen exposure, irritant exposure resulted in cells within the dermis becoming CD1a{sup −}/CD14{sup +}/CD68{sup +} which is characteristic of a phenotypic switch of MUTZ-LC to a macrophage-like cell in the dermis. This phenotypic switch was blocked with anti-IL-10. Mechanisms previously identified as being involved in LC activation and migration in native human skin could thus be reproduced in the in vitro constructed skin equivalent model containing functional LC. This model therefore provides a unique and relevant research tool to study human LC biology in situ under controlled in vitro conditions, and will provide a powerful tool for hazard identification, testing novel therapeutics and identifying new drug targets. - Highlights: • MUTZ-3 derived Langerhans cells integrated into skin equivalents are fully functional. • Anti-CXCL12 blocks allergen-induced MUTZ-LC migration. • Anti-CCL5 blocks irritant-induced MUTZ-LC migration. • Irritant mediated MUTZ-LC trans-differentiation to macrophage-like cell in dermis. • Trans-differentiation of MUTZ-LC is IL-10 dependent.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2014

Sci—Fri PM: Topics — 04: What if bystander effects influence cell kill within a target volume? Potential consequences of dose heterogeneity on TCP and EUD on intermediate risk prostate patients

In vitro evidence has suggested that radiation induced bystander effects may enhance non-local cell killing which may influence radiotherapy treatment planning paradigms. This work applies a bystander effect model, which has been derived from published in vitro data, to calculate equivalent uniform dose (EUD) and tumour control probability (TCP) and compare them with predictions from standard linear quadratic (LQ) models that assume a response due only to local absorbed dose. Comparisons between the models were made under increasing dose heterogeneity scenarios. Dose throughout the CTV was modeled with normal distributions, where the degree of heterogeneity was then dictated by changing the standard deviation (SD). The broad assumptions applied in the bystander effect model are intended to place an upper limit on the extent of the results in a clinical context. The bystander model suggests a moderate degree of dose heterogeneity yields as good or better outcome compared to a uniform dose in terms of EUD and TCP. Intermediate risk prostate prescriptions of 78 Gy over 39 fractions had maximum EUD and TCP values at SD of around 5Gy. The plots only dropped below the uniform dose values for SD ∼ 10 Gy, almost 13% of the prescribed dose. The bystander model demonstrates the potential to deviate from the common local LQ model predictions as dose heterogeneity through a prostate CTV is varies. The results suggest the potential for allowing some degree of dose heterogeneity within a CTV, although further investigations of the assumptions of the bystander model are warranted.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2023

Natural Language Processing for Text Based Event Extraction: Identifying Events of Interest Related to Worldwide State-Sponsored Civil Nuclear Power

Beginning in FY20, SRNL was funded by the National Nuclear Security Administration’s Office of Defense Nuclear Non-Proliferation Research and Development to develop a prototype natural language processing/natural language understating machine learning-based modeling and analysis pipeline to extract and forecast events of interest from massive open data sources. The working hypothesis within the approach is that contextual shifts in key words and phrases act as indicators of events of interest over time. Therefore, by identifying points in time where contextual shifts occur, events of interest can be extracted along with explicit and implicit connections of entities and activities. The development of the preliminary prototype pipeline proved successful, meriting further testing of the pipeline on more broad topical domains and in a worldwide data environment. Therefore, SRNL, in collaboration with the Sanghani Center for Artificial Intelligence and Data Analytics at Virginia Tech, have continued development with a test case of identifying events of interest related to worldwide state-sponsored civil nuclear power in open data sources. In the first year of this follow-on effort, the team has curated domain-specific data corpuses using an automated scheme and applied the modeling and analysis pipeline. This robust, focused, and efficient approach consists of an ensemble of analyses applied to time dependent word embedding models that are trained on the data corpuses. In this report, the team has demonstrated the capability of the existing pipeline (as development has continued in parallel) by exploring several specific case-studies centered around Rosatom’s international activities regarding the planning, construction, operation, and/or shutdown of nuclear reactors. A basic timeline events has been generated by manually cataloging known “milestone” events that have occurred at reactors in Turkey, Finland, Hungary, and Egypt and compared with the output of the modeling pipeline. In this approach, the team has characterized the lead time using the prototype pipeline, as well as the ability to capture relevant information, which proved 100% successful. A deep dive example of the Akkuyu reactor (Turkey) is presented that shows the breadth of information that can be captured using the approach. In this case study, events were extracted pertaining to the planning/construction of Akkuyu including protests from the population, information campaigns in response to the protests, forged regulatory documents and lawsuits, budgetary/shareholder information, geopolitical tensions, and the various construction milestones. This has demonstrated the pipeline’s utility as a research aid or real-time event extraction tool, where summary-level information and detailed text extractions from millions of articles or Tweets across long time periods can be generated with significantly less effort than current techniques.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗