Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Deep Direct Use”

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 631 records · Page 35
All sourcesNASA NTRSDOE OSTI
DOE OSTI2026

Tuning the Solvation and Solubility Properties of Molecularly Heterogeneous Nonionic Deep Eutectic Solvents via Interface Organization

Common separation techniques, such as liquid− liquid extraction, are usually used for extractions and purifications due to their industrial scalability and affordability. However, these well-established practices are hindered by low selectivity and challenges in recovering solutes and solvents. Deep eutectic solvents (DES), a fairly new type of solvent, have the potential to overcome these issues. DESs are binary mixtures whose physical properties can be tuned by selecting the appropriate precursors to facilitate and/or enhance processes such as extraction. A promising DES for selective separations is formed when lauric acid (LA) is mixed with N-methylacetamide (NMA). This LA-NMA DES has a heterogeneous microscopic structure that can solvate compounds with completely different polarities. This study explores how forming an organized structure on the mesoscale affects the solubility of nonpolar solutes in nonionic DESs. To this end, the molecular and mesoscale structures and their effect on the solubility and solvation properties are evaluated for the LA-NMA DES and two new DESs with slight chemical variations in their precursors. It is observed that the organization of the nonpolar DES domains and, consequently, of their interfaces directly relates to the solubility of nonpolar compounds. Specifically, correctly selecting the DES precursors that form organized nonpolar domains leads to an organized interface in which the nonpolar solutes are solvated, thereby increasing the solubility. Additionally, enhanced dissolution power was observed in a completely different DES with mesoscale order in its molecular structure and composed of menthol and lauric acid. The latter result further validates the proposed tunability of the DES dissolution power through organized interfaces, extending it beyond a specific DES family and opening the possibility of new extraction-tailored designer solvents.

Extraction↗
DOE OSTI2024

High-throughput chemical imaging for optimizing biofuel synthesis using synthetic biology (Final Technical Report)

Fatty acids can be produced biosynthetically in microbes and these compounds can serve as precursors to biodiesels and other high value oleochemicals. However, progress on engineering fatty acid biosynthesis, and biofuel synthesis more generally, has been hindered by current quantification methods that are either indirect or not amenable to high-throughput or single-cell resolution screening. In this project, we assembled an interdisciplinary team with complimentary expertise in synthetic biology and microscopy, metabolic engineering, and chemical imaging to address these challenges. We used chemical imaging to directly measure lipid biosynthesis in Escherichia coli engineered to produce fatty acids, obtaining detailed single-cell resolution measurements. We deployed stimulated Raman scattering (SRS) microscopy in concert with multiplexed genome engineering and gene circuit design strategies from synthetic biology to optimize production of fatty acids. These results provided novel insight into cell-to-cell heterogeneity present in biofuel production strains. In addition, we introduced new chemical imaging methods which are label-free and do not require fluorescent reporters. These efforts were complemented by other studied developing foundational tools for regulation and control, which offer excellent potential for advancing researchers’ ability to rapidly design, build, and test strains for enhanced biofuel synthesis.

09 BIOMASS FUELS↗
DOE OSTI2022

J-PLUS: Uncovering a large population of extreme [OIII] emitters in the local Universe

Over the past decades, several studies have discovered a population of galaxies that undergo very strong star formation events. They are called extreme emission line galaxies (EELGs). We exploit the capabilities of the Javalambre Photometric Local Universe Survey (J-PLUS), a wide-field multifilter survey, with which 2000 square degrees of the northern sky are already observed. We use it to identify EELGs at low redshift by their [OIII]5007 emission line. We intend to provide a more complete, deep, and less biased sample of local EELGs. We selected objects with an excess of flux in the J-PLUS medium-band J0515 filter, which covers the [OIII] line at z < 0.06. We removed contaminants (stars and higher-redshift systems) using J-PLUS and WISE infrared photometry, with SDSS spectra as a benchmark. We performed spectral energy distribution fitting to estimate the physical properties of the galaxies: line fluxes, equivalent widths (EWs), masses, stellar population ages, and so on. We identify 466 EELGs at z < 0.06 with [OIII] EW over 300 Å and an r-band magnitude below 20, of which 411 were previously unknown. Most show compact morphologies, low stellar masses (log(M • /M ⊙ ) ~ 8.13 –0.58 +0.61 ), low dust extinction (E( B–V ) ~ 0.1 –0.1 +0.2 ), and very young bursts of star formation (3.0 –2.0 +2.7 Myr). Our method is up to ~20 times more efficient in detecting EELGs per Mpc 3 than broadband surveys, and it is as complete as magnitude-limited spectroscopic surveys (but reaches fainter objects). The sample is not directly biased against strong Hα emitters, in contrast with works using broadband surveys. We demonstrate that J-PLUS can identify a large sample of previously unknown EELGs showing unique properties following a clear selection process. A fraction of the EELGs are probably similar to the first galaxies in the Universe, but they are at a much lower redshift, which makes them ideal targets for follow-up studies.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2026

Detecting and Characterizing Fracture Zones Using a Convolutional Neural Network

This project directly supports the Geothermal Technologies Office (GTO) objectives outlined in the Multi-Year Program Plan (MYPP) by advancing two key research areas: “Exploration and Characterization” and “Data, Modeling, and Analysis.” This project has successfully demonstrated a pre-drilling ability to image and characterize the distribution and connectivity of subsurface faults and fractures, key parameters for identifying permeable pathways that enable geothermal fluids to circulate and produce energy. Specifically, we developed and implemented innovative machine learning methodologies to enhance geothermal exploration. Large-scale faults were detected using a Convolutional Neural Network (CNN), while small-scale fractures were characterized using a novel Double-Beam Neural Network (DBNN). These tools have proven both technically effective and cost-efficient by reducing reliance on expensive exploratory drilling. Through collaboration with our geothermal industry partner, this research has significantly advanced techniques for identifying hidden geothermal systems and extending the productive lifespan of existing geothermal fields. We applied our methods to two geothermal fields—Soda Lake (Nevada) and Lightning Dock (New Mexico)—to identify shallow steam-charged fracture zones and characterize deep faults at depths of 1.5-2 km. The steam zone identified at the Soda Lake geothermal field showed excellent agreement with prior drilling data, validating the effectiveness of our approaches. In addition, the analysis revealed three new prospective drilling targets for further development and verification. The outcomes of this project improve our scientific understanding of geothermal reservoir behavior, enhance exploration efficiency, extend the economic life of existing geothermal plants. Ultimately, these advancements contribute to GTO’s goal of achieving more sustainable, affordable, and data-driven geothermal energy development across the United States.

15 GEOTHERMAL ENERGY↗
DOE OSTI2025

Advancing the Frontiers of Deep Learning for Low-Dose 3D Cone-Beam CT Reconstruction

X-ray computed tomography (CT) is an important noninvasive medical imaging modality for studying the structural details of internal organs. Image reconstruction in CT is an inverse problem of recovering an object's internal structure from the absorption profile of X-ray beams (sinogram) measured using a detector. The classical variational approach for CT reconstruction minimizes an energy functional using an appropriate iterative algorithm. Motivated by the success of deep learning (DL), researchers have begun to leverage training data and enhanced computing capabilities in recent years to produce high-fidelity reconstructed images. Nonetheless, much of the academic research in DL algorithms for CT has focused primarily on the two-dimensional setting (with simplified forward operators and noise model) for proofs-of-concept, and a comprehensive benchmarking of various classical and data-driven CT reconstruction approaches has not beenundertaken. The key objective of our CT reconstruction grand challenge was to promote methodological advancements for both classical and DL-based approaches for clinical CT with a reasonably accurately simulated 3D CT forward operator and noise model. We have utilized the publicly available LIDC-IDRI dataset and simulated sinograms and FDK images corresponding to two dose levels (clinical- and low-dose, constituting two tracks of the challenge) starting from the normal-dose images as the ground truth. In this paper, we summarize the motivation, context, and results of our challenge, and highlight the future research directions in DL for clinical CT.

X-ray tomography↗
DOE OSTI2026

Xanthos-Lake Model Source Code

This repository contains the source code for Xanthos-Lake, a lake-modeling extension of the Xanthos framework that introduces a coupled lake component comprising the Xanthos-Lake Snow and Ice Model (xLSIM) and the Xanthos-Lake Water Balance Model (xLWBM). xLSIM is a basin-aware machine-learning model for lake snow, ice, and thermal conditions. It predicts monthly lake ice thickness, snow depth, snow-cover fraction, mixing-layer temperature, and lake ice fraction from meteorological forcing and lake surface-area information. It uses sequence-based deep-learning architectures, including Transformer and hybrid Long Short-Term Memory–Transformer (LSTM–Transformer) models, together with seasonal encoding, multi-lake learning, physical masking, and basin-level cryospheric and non-cryospheric classification. The training workflow uses Ray for scalable execution and includes optional Ray Tune hyperparameter optimization. Model predictions, observations, diagnostics, and feature-importance outputs are written in NetCDF. xLWBM is the water-balance component of the new lake framework. It simulates monthly lake storage, surface area, evaporation, inflow, outflow, and lake–groundwater exchange. It combines physical water-balance equations with calibrated bathymetric relationships, weir-based outlet flow, modified Penman open-water evaporation, groundwater head relaxation, Penman–Monteith snow and ice sublimation, and snow, ice, and thermal conditions supplied by xLSIM. The model calibrates lake parameters against satellite-derived surface-area data, using evaporation-based calibration where surface-area data are unavailable, and supports small, medium, and large lake classes. For large lakes, xLWBM is integrated with the managed-routing workflow so that lake storage and outflow interact directly with downstream river routing and reservoir operations. Together, xLSIM and xLWBM provide Xanthos with a coupled lake-modeling capability. xLSIM supplies the snow, ice, and thermal conditions that affect lake evaporation and snow- and ice-related water exchanges, while xLWBM translates those conditions into dynamic lake storage, surface area, evaporation, and discharge. In return, xLWBM supplies evolving lake surface area to xLSIM. This coupling enables Xanthos to represent lakes as active hydrologic components within basin-scale water-availability and routing simulations.

Machine Learning↗
DOE OSTI2026

Machine Learning-Based Anomaly Detection for PMT Data Quality Monitoring in the SBN and DUNE

Maintaining high-quality detector data is essential for achieving the scientific objectives of the Short-Baseline Neutrino (SBN) Program at Fermilab. Current data quality monitoring (DQM) procedures rely primarily on threshold-based metrics and manual inspection of detector monitoring plots, making the detection of subtle or gradually developing anomalies both time-consuming and dependent on expert interpretation. This project developed and evaluated a machine-learning workflow for automatically identifying anomalous photomultiplier tube (PMT) channels in the Short-Baseline Near Detector (SBND) using optical-hit amplitude data. A Python-based analysis program was developed to process ROOT files, extract statistical features describing individual PMT amplitude distributions, and generate feature vectors for anomaly detection. These features were used to train an Isolation Forest model using data representing normal detector operation. The trained model was subsequently applied to independent detector runs to identify channels exhibiting statistically unusual behavior relative to the learned reference response. To support expert interpretation, the workflow generated complementary diagnostic products, including anomaly score distributions, normalized amplitude comparisons, decision-tree visualizations, and principal component analysis (PCA) projections. This project demonstrated the feasibility of integrating unsupervised machine learning into detector data-quality monitoring and developed a complete workflow for automated PMT performance assessment to aid expert-driven review. Beyond its technical contributions, the VFP appointment fostered a research collaboration between Aurora University and Fermilab and provided direct workforce development benefits by training the visiting faculty member in detector-scale machine-learning methods that are now being incorporated into undergraduate coursework and research. The methodology developed here provides a foundation for future applications to ProtoDUNE and other liquid argon time projection chamber (LArTPC) detectors, contributing to ongoing efforts to improve detector reliability, reduce manual monitoring requirements, and enable scalable data quality monitoring for future large-scale neutrino experiments, including the Deep Underground Neutrino Experiment (DUNE).

Colón Santana, Juan A. [Unlisted, US, IL]↗
DOE OSTI2025

Digital Real-Time Simulation and Power Quality Analysis of a Hydrogen-Generating Nuclear-Renewable Integrated Energy System

This paper investigates the challenges and solutions associated with integrating a hydrogen-generating nuclear-renewable integrated energy system (NR-IES) under a transactive energy framework. The proposed system directs excess nuclear power to hydrogen production during periods of low grid demand while utilizing renewables to maintain grid stability. Using digital real-time simulation (DRTS) in the Typhoon HIL 404 model, the dynamic interactions between nuclear power plants, electrolyzers, and power grids are analyzed to mitigate issues such as harmonic distortion, power quality degradation, and low power factor caused by large non-linear loads. A three-phase power conversion system is modeled using the Typhoon HIL 404 model and includes a generator, a variable load, an electrolyzer, and power filters. Active harmonic filters (AHFs) and hybrid active power filters (HAPFs) are implemented to address harmonic mitigation and reactive power compensation. The results reveal that the HAPF topology effectively balances cost efficiency and performance and significantly reduces active filter current requirements compared to AHF-only systems. During maximum electrolyzer operation at 4 MW, the grid frequency dropped below 59.3 Hz without filtering; however, the implementation of power filters successfully restored the frequency to 59.9 Hz, demonstrating its effectiveness in maintaining grid stability. Future work will focus on integrating a deep reinforcement learning (DRL) framework with real-time simulation and optimizing real-time power dispatch, thus enabling a scalable, efficient NR-IES for sustainable energy markets.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗
DOE OSTI2023

Small $x$ physics beyond eikonal approximation: an effective Hamiltonian approach

Understanding the spin structure of hadrons in the small x regime is an important direction to unravel the spin puzzle in hadronic physics. To include spin degrees of freedom in the small x regime requires going beyond the usual eikonal approximation in high energy QCD. We developed an effective Hamiltonian approach to study spin related observables in the small x regime using the shockwave formalism. The small-x effective Hamiltonian incorporates both quark and gluon propagators in the background fields and the background field induced interaction vertices up to next-to-eikonal order. A novel feature of sub-eikonal interactions is the background gluon field induced gluon radiation inside the shockwave. Its relation to chromo-electrically polarized Wilson line correlator is established both in small x helicity evolution and in longitudinal double-spin asymmetry for gluon production.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2021

Charge Transport and Space-Charge Formation in Cd 1 - x Zn x Te 1 - y Se y Radiation Detectors

The electron- and hole-transport properties in cadmium zinc telluride selenide (CZTS) crystals are studied using a laser-induced transient-current technique with pulsed and dc bias. The internal electric field profile and velocity of surface recombination are determined by Monte Carlo simulations of electron and hole transient currents combined with a numerical solution of the drift-diffusion equation coupled with Poisson’s equation. Electron and hole drift mobilities of μe = 830 cm 2 /Vs and μh = 40 cm 2 /Vs, respectively, are determined. We also develop a simple technique for evaluating surface recombination directly from measured current waveforms without the need for numerical simulation. The good quality of the prepared detector at pulsed bias, with electron- and hole-mobility-lifetime products of (μτ)e = 1.9 × 10 -3 cm 2 /V and (μτ)h = 1.4 × 10 -4 cm 2 /V, respectively, are observed. The formation of a positive space charge, originating from hole injection combined with a recombination level, is found. We observe a significant position dependence of the lifetime of electrons and holes in dc bias due to hole injection. The experiment is successfully fitted by a simple model dominated by a single deep recombination level with an energy of E t =E C -0.73eV; concentration of 7.3 × 10 11 cm -3 ; and electron- and hole-capture cross sections of 3.5 × 10 -14 cm 2 and 6.5 × 10 -14 cm 2 , respectively.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗
DOE OSTI2026

Development of a deep potential model for F and CF 2 etching of Si and SiO 2

An understanding of plasma-surface interactions at increasingly smaller scales is invaluable for the development of novel technologies and processing techniques. Molecular dynamics (MD) simulations can provide insights into atomic-scale interactions, though they are restricted by the availability of interatomic potentials. Machine learning methods, such as Deep Potential Molecular Dynamics (DeepMD), provide a systematic framework for the development of accurate and flexible ab initio-based models. In this work, we develop DeepMD models for the ion-enhanced etching of Si and SiO 2 by F and CF 2 radicals. We employ an active learning process to expand the data set on which the model is trained and demonstrate its effect on the model accuracy. The DeepMD results are compared to data from classical MD simulations and experiments. Physical sputtering yields of SiO 2 by Ar + ions show good agreement with previous simulation results using conventional interatomic potentials, though the predicted depth profiles are different. Etching yields are calculated as a function of ion energy and neutral to ion flux ratio for the Ar + ion-enhanced etching of SiO 2 and Si by F atoms, as well as for etching of SiO 2 by CF 2 radicals, showing reasonable agreement with experimental data. Finally, an ion-enhanced surface kinetic model is fitted to the DeepMD etch yields, and the fitted parameters are compared to quantities computed directly from DeepMD simulations. This study illustrates how molecular dynamics simulations using machine learning potentials can provide an accurate model of etching processes relevant to device manufacturing.

Kounis-Melas, Andreas [Princeton Univ., NJ (United↗
DOE OSTI2021

Deep Learning Approaches to Surrogates for Solving the Diffusion Equation for Mechanistic Real-World Simulations

In many mechanistic medical, biological, physical, and engineered spatiotemporal dynamic models the numerical solution of partial differential equations (PDEs), especially for diffusion, fluid flow and mechanical relaxation, can make simulations impractically slow. Biological models of tissues and organs often require the simultaneous calculation of the spatial variation of concentration of dozens of diffusing chemical species. One clinical example where rapid calculation of a diffusing field is of use is the estimation of oxygen gradients in the retina, based on imaging of the retinal vasculature, to guide surgical interventions in diabetic retinopathy. Furthermore, the ability to predict blood perfusion and oxygenation may one day guide clinical interventions in diverse settings, i.e., from stent placement in treating heart disease to BOLD fMRI interpretation in evaluating cognitive function (Xie et al., 2019; Lee et al., 2020). Since the quasi-steady-state solutions required for fast-diffusing chemical species like oxygen are particularly computationally costly, we consider the use of a neural network to provide an approximate solution to the steady-state diffusion equation. Machine learning surrogates, neural networks trained to provide approximate solutions to such complicated numerical problems, can often provide speed-ups of several orders of magnitude compared to direct calculation. Surrogates of PDEs could enable use of larger and more detailed models than are possible with direct calculation and can make including such simulations in real-time or near-real time workflows practical. Creating a surrogate requires running the direct calculation tens of thousands of times to generate training data and then training the neural network, both of which are computationally expensive. Often the practical applications of such models require thousands to millions of replica simulations, for example for parameter identification and uncertainty quantification, each of which gains speed from surrogate use and rapidly recovers the up-front costs of surrogate generation. We use a Convolutional Neural Network to approximate the stationary solution to the diffusion equation in the case of two equal-diameter, circular, constant-value sources located at random positions in a two-dimensional square domain with absorbing boundary conditions. Such a configuration caricatures the chemical concentration field of a fast-diffusing species like oxygen in a tissue with two parallel blood vessels in a cross section perpendicular to the two blood vessels. To improve convergence during training, we apply a training approach that uses roll-back to reject stochastic changes to the network that increase the loss function. The trained neural network approximation is about 1000 times faster than the direct calculation for individual replicas. Because different applications will have different criteria for acceptable approximation accuracy, we discuss a variety of loss functions and accuracy estimators that can help select the best network for a particular application. We briefly discuss some of the issues we encountered with overfitting, mismapping of the field values and the geometrical conditions that lead to large absolute and relative errors in the approximate solution.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2020

The Role of Ice Splintering on Microphysics of Deep Convective Clouds Forming Under Different Aerosol Conditions: Simulations Using the Model With Spectral Bin Microphysics

Observations during the Ice in Clouds Experiment-Tropical (ICE-T) field experiment show that the ice particles concentration in a developing deep convective clouds at the level of T = -15 °C reached about 500 L -1 , that is, many orders higher than that of ice-nucleating particle. To simulate microphysics of these clouds, the 2-D Hebrew University Cloud model (HUCM) with spectral bin microphysics was used in which two main types of ice multiplication mechanisms were included in addition to the Hallet-Mossop mechanism. In the first ice multiplication mechanism ice splinters form by drop freezing and drop-ice collisions. Ice multiplication of this type dominates during developing stage of cloud evolution, when liquid water content is significant. At later stage when clouds become nearly glaciated, ice crystals are produced largely by ice splintering during ice-ice collisions (the second ice multiplication mechanism). Simulations show that droplet size distributions, as well as size distributions of ice particles, agree well with the measurements during ICE-T. Simulations with different cloud condensation nuclei concentrations show the existence of the “optimum” cloud condensation nuclei concentration (or droplet concentration), at which concentration of ice splinters reaches maximum. In these simulations ice nucleation caused by the direct formation of ice crystals upon ice-nucleating particles, as well as the Hallett-Mossop process, has a negligible contribution to the ice crystal concentration.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2021

Ultra-high energy density relativistic plasmas from nanostructures: scaling to ultra-high intensities (Final Report)

This project investigated the ultra-high energy density (UHED) regime of matter found in the center of stars, using a compact PW-class laser. These extreme conditions are typically only obtained in the laboratory in the central hot-spot of spherically imploded capsules in inertial confinement fusion experiments driven by the world’s largest lasers. We have shown that near-solid density arrays of aligned nanostructures can be volumetrically heated to multi-keV temperatures by irradiation at relativistic intensity with ultrafast laser pulses of modest energy, opening a path for the generation of UHED plasmas with compact lasers. This new UHED plasma generation approach promises to create an environment with extreme energy densities and degrees of ionization, record conversion of optical laser light into ultrafast x-ray pulses, gigantic magnetic fields and pressures, and directed beams of high energy particles . In this project we achieved record degree of ionization in volumetric heating solid density and near-solid density plasmas. Gold plasmas which spectra is characterized by L shell transition emission from ions with charge up to the Ne-like state, Au+ 72 were generated using ultrafast laser pulses of less than 10 J of energy from a compact laser focused to an intensity of ~ 3x10 21 Wcm -2 . We also conducted measurements to determine the heat penetration depth in Ni nanowire arrays as compared to Ni foil targets by monitoring the line emission of a Co buried tracer underneath a variable amount of Ni. The measurements revealed that the nanowire plasmas are roughly six times larger in depth than the solid density target plasmas. A result of this increased plasma in nanowire arrays is a greatly increased conversion of optical laser light into > 1 KeV x-rays, a record conversion efficiency of 20 %. Critical to the realization of the proposed experiments was the generation of ultrafast laser pulses with ultra-high contrast that can deposit the energy deep into the nanowire arrays before the nanowires explode to create a continuous plasma. Supporting the proposed experiments was the recent demonstration at Colorado State University of a Petawatt-class laser that emits 30 fs pulses at high repetition rates. The experiments combined this unique laser tool with a variety of tailored nanowire arrays fabricated in house and with an extensive suite of diagnostics. The experiments were accompanied by 3-dimensional particle-in-cell simulations and detailed atomic physics simulations with transient kinetics and radiation transport. The proposed research allowed us to continue training Ph.D students and post-docs with broad experience in HEDP Science.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2020

Effects of cation stoichiometry on surface morphology and crystallinity of ZnGeN 2 films grown on GaN by metalorganic chemical vapor deposition

ZnGeN 2 films were grown on GaN-on-sapphire templates via metalorganic chemical vapor deposition. Energy dispersive X-ray spectroscopy was used to estimate the Zn/(Zn+Ge) composition ratio in the films. This ratio decreased with increase in growth temperature but increased with increase in total reactor pressure or Zn/Ge precursor flow rate ratio. Systematic mapping of these key growth parameters has allowed us to identify the growth window to achieve ZnGeN 2 with stoichiometric cation composition. Compositional and statistical analyses performed on data acquired from atom probe tomography provided insight into the local compositional homogeneity. The cations Zn and Ge did not demonstrate segregation or clustering at the sub-nanometer level. Based on X-ray diffraction 2θ-ω scan profiles and transmission electron microscope nano-diffraction patterns, the films with near-stoichiometric cation ratios were single crystalline with planar surfaces, whereas zinc-rich or zinc-poor films were polycrystalline with nonplanar surfaces. The growth direction of the single crystalline ZnGeN 2 films on GaN templates was along the c-axis. Room temperature Raman spectra showed features associated with the phonon density of states, indicating the presence of cation disorder in the lattice. A cathodoluminescence peak associated with transitions involving deep level defects was observed around 640 nm. The intensity of this peak increased by almost 2.5 times as the temperature was reduced to 77 K from room temperature. A similar peak was observed in photoluminescence spectra collected at 80 K.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2020

Variations of Stress Parameters in the Southern California Plate Boundary Around the South Central Transverse Ranges

We examine the stress field in Southern California with a focus on the region near the South Central Transverse Ranges (SCTR), using a refined stress inversion methodology. We independently invert declustered and aftershocks focal mechanisms from the years 1981 to 2017. Comparisons between the stress field orientations and the stress ratios from these inversions provide information on local sources of the stress field. Regionally, the SHmax tends to be toward NNE and the stress ratios vary from transtensional stress regime near the Eastern California Shear Zone (ECSZ), to strike-slip faulting near the SCTR, and toward transpression near the Western Transverse Ranges. Detailed analysis of stress parameters near the SCTR indicates deviations from the regional strike-slip faulting including the following: (1) The San Bernardino Mountain area shows S Hmax direction toward NNW and transpressional stress components likely associated with the motions of the San Andreas Fault and ECSZ; (2) the Cajon Pass and San Gorgonio Pass show transpressional stress regime near the bottom of the seismogenic zone likely associated with the elevated topography; and (3) in Crafton Hills, rotation of the principal stress plunges and SHmax direction and transtensional stress regime below ~10 km, along with lower estimated apparent friction coefficient, suggest a weak fault possibly associated with deep creep. The performed multiscale analysis resolves effects of regional and local loadings. The average stress parameters in the analyzed 37 years do not show significant temporal variations of stress field near the SCTR region.

58 GEOSCIENCES↗
DOE OSTI2023

Scaling Resolution of Gigapixel Whole Slide Images Using Spatial Decomposition on Convolutional Neural Networks

Gigapixel images are prevalent in scientific domains ranging from remote sensing, and satellite imagery to microscopy, etc. However, training a deep learning model at the natural resolution of those images has been a challenge in terms of both, overcoming the resource limit (e.g. HBM memory constraints), as well as scaling up to a large number of GPUs. In this paper, we trained Residual neural Networks (ResNet) on 22,528 x 22,528-pixel size images using a distributed spatial decomposition method on 2,304 GPUs on the Summit Supercomputer. We applied our method on a Whole Slide Imaging (WSI) dataset from The Cancer Genome Atlas (TCGA) database. WSI images can be in the size of 100,000 x 100,000 pixels or even larger, and in this work we studied the effect of image resolution on a classification task, while achieving state-of-the-art AUC scores. Moreover, our approach doesn't need pixel-level labels, since we're avoiding patching from the WSI images completely, while adding the capability of training arbitrary large-size images. This is achieved through a distributed spatial decomposition method, by leveraging the non-block fat-tree interconnect network of the Summit architecture, which enabled GPU-to-GPU direct communication. Finally, detailed performance analysis results are shown, as well as a comparison with a data-parallel approach when possible.

Tsaris, Aristeidis (aris)↗
DOE OSTI2023

BCG alignment with the locations of cluster members and the large-scale structure out to 10 R 200

Using a sample of >200 clusters, each with typically 100–200 spectroscopically confirmed cluster members, we search for a signal of alignment between the position angle (PA) of the Brightest Cluster Galaxy (BCG) and the distribution of cluster members on the sky about the cluster centre out to projected distances of 3 R 200 . The deep spectroscopy, combined with corrections for spectroscopic incompleteness, makes our sample ideal to determine alignment signal strengths. We also use an SDSS based skeleton of the filamentary large-scale structure (LSS), and measure BCG alignment with the location of the LSS skeleton segments on the sky out to projected distances of 10 R 200 . The alignment signal is measured using three separate statistical measures; Rao’s spacing test (U), Kuiper’s V parameter (V), and the Binomial probability test (P). The significance of the BCG alignment signal with both cluster members and LSS segments is extremely high (1 in a million chance or less to be drawn randomly from a uniform distribution). We investigate a wide set of parameters that may influence the strength of the alignment signal. Clusters with more elliptical-shaped BCGs show stronger alignment with both their cluster members and LSS segments. Also, selecting clusters with closely connected filaments, or using a luminosity-weighted LSS skeleton, increases the alignment signal significantly. Alignment strength decreases with increasing projected distance. Combined, these results provide strong evidence for the growth of clusters and their BCGs by preferential feeding along the direction of the filaments in which they are embedded.

79 ASTRONOMY AND ASTROPHYSICS↗