Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “deep well”

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 847 records · Page 47
All sourcesNASA NTRSDOE OSTI
DOE OSTI2023

Towards inverse microstructure-centered materials design using generative phase-field modeling and deep variational autoencoders

The field of Integrated Computational Materials Engineering (ICME) combines a broad range of methods to study materials’ responses over a spectrum of length scales. A relatively unexplored aspect of microstructure-sensitive materials design is uncertainty propagation and quantification (UP/UQ) of materials’ microstructure, as well as establishing process-structure–property (PSP) relationships for inverse material design. In this study, an efficient UP technique built on the idea of changing probability measures and a deep generative unsupervised representative machine learning method for microstructure-based design of thermal conductivity of materials is proposed. Probability measures are used to represent microstructure space, and Wasserstein metrics are used to test the efficiency of the UP method. By using deep Variational AutoEncoder (VAE), we identify the correlations between the material/process parameters and the thermal conductivity of heterogeneous dual-phase microstructures. Through high-throughput screening, UP, and the deep-generative VAE method, PSP relationships that are too complex can be revealed by exploiting the materials’ design space with an emphasis on microstructures. As a last point, we demonstrate generative machine learning serves as a useful tool for inverse microstructure-centered materials design, and we demonstrate this by examining the inverse design of thermal conductivity in nano-structured materials. Here, the results reveal the effects of morphology, volume fraction, characteristic length scale, and the individual thermal diffusivity of phases on the thermal conductivity of dual-phase alloys. Our findings emphasize the advantages of high-throughput phase-field modeling and generative deep learning for linking PSP and inverse microstructure-centered materials design.

36 MATERIALS SCIENCE↗
NASA NTRS2016

Filter Efficiency and Leak Testing of Returned ISS Bacterial Filter Elements After 2.5 Years of Continuous Operation

The atmosphere revitalization equipment aboard the International Space Station (ISS) and future deep space exploration vehicles provides the vital functions of maintaining a habitable environment for the crew as well as protecting the hardware from fouling by suspended particulate matter. Providing these functions are challenging in pressurized spacecraft cabins because no outside air ventilation is possible and a larger particulate load is imposed on the filtration system due to lack of sedimentation in reduced gravity conditions. The ISS Environmental Control and Life Support (ECLS) system architecture in the U.S. Segment uses a distributed particulate filtration approach consisting of traditional High-Efficiency Particulate Adsorption (HEPA) filters deployed at multiple locations in each module. These filters are referred to as Bacteria Filter Elements (BFEs). As more experience has been gained with ISS operations, the BFE service life, which was initially one year, has been extended to two to five years, dependent on the location in the U.S. Segment. In previous work we developed a test facility and test protocol for leak testing the ISS BFEs. For this work, we present results of leak testing a sample set of returned BFEs with a service life of 2.5 years, along with particulate removal efficiency and pressure drop measurements. The results can potentially be utilized by the ISS Program to ascertain whether the present replacement interval can be maintained or extended to balance the on-ground filter inventory with extension of the lifetime of ISS to 2024. These results can also provide meaningful guidance for particulate filter designs under consideration for future deep space exploration missions.

ISS Air Quality↗
NASA NTRS2000

The Deep Space 1 and Space Technology 4/Champollion Missions

NASA's New Millennium Program (NMP) is designed to develop, test, and flight validate new, advanced technologies for planetary and Earth exploration missions, using a series of low cost spacecraft. Two of NMP's current missions include encounters with comets and asteroids. The Deep Space 1 mission was launched on October 24, 1998 and will fly by asteroid 1992 KD on July 29, 1999, and possibly Comet Wilson-Harrington and/or Comet Borrelly in 2001. The Space Technology 4/Champollion mission will be launched in April, 2003 and will rendezvous with, orbit and land on periodic Comet Tempel 1 in 2006. ST-4/Champollion is a joint project with CNES, the French space agency. The DS-1 mission is going well since launch and has already validated several major technologies, including solar electric propulsion (SEP), solar concentrator arrays, a small deep space transponder, and autonomous navigation. The spacecraft carries two scientific instruments: MICAS, a combined visible camera and UV and IR spectrometers, and PEPE, an ion and electron spectrometer. Testing of the science instruments is ongoing. Following the asteroid encounter in July, 1999, DS-1 will go on to encounters with one or both comets if NASA approves funding for an extended mission. The ST-4/Champollion mission will use an advanced, multi-engine SEP system to effect a rendezvous with Comet P/Tempel 1 in February, 2006, after a flight time of 2.8 years. After orbiting the comet for several months in order to map its surface and determine its gravity field, ST-4/Champollion will descend to the comet's surface and will anchor itself with a 3-meter long harpoon. Scientific experiments include narrow and wide angle cameras for orbital mapping, panoramic and near-field cameras for landing site mapping, a gas chromatograph/mass spectrometer, a combined microscope and infrared spectrometer, and physical properties probes. Cometary samples will be obtained from depths up to 1.4 meters. The spacecraft is solar powered with rechargeable batteries, thus allowing a long duration mission on the nucleus surface. At the time of this writing, the ST-4/Champollion spacecraft was undergoing a major redesign to fit within NASA cost constraints, and approval of the mission is pending.

Weissman, Paul R.↗
NASA NTRS2000

Comet and Asteroid Missions in NASA's New Millennium Program

NASA's New Millennium Program (NMP) is designed to develop, test, and flight validate new, advanced technologies for planetary and Earth exploration missions, using a series of low cost spacecraft. Two of NMP's current missions include encounters with comets and asteroids. The Deep Space 1 mission was launched on October 24, 1998 and will fly by asteroid 1992 KD on July 29, 1999, and possibly Comet Wilson-Harrington and/or Comet Borrelly in 2001. The Space Technology 4/Champollion mission will be launched in April, 2003 and will rendezvous with, orbit and land on periodic Comet Tempel 1 in 2006. ST-4/Champollion is a joint project with CNES, the French space agency. The DS-1 mission is going well since launch and has already validated several major technologies, including solar electric propulsion (SEP), solar concentrator arrays, a small deep space transponder, and autonomous navigation. The spacecraft carries two scientific instruments: MICAS, a combined visible camera and UV and IR spectrometers, and PEPE, an ion and electron spectrometer. Testing of the science instruments is ongoing. Following the asteroid encounter in July, 1999, DS-1 will go on to encounters with one or both comets if NASA approves funding for an extended mission. The ST-4/Champollion mission will use an advanced, multi-engine SEP system to effect a rendezvous with Comet P/Tempel 1 in February, 2006, after a flight time of 2.8 years. After orbiting the comet for several months in order to map its surface and determine its gravity field, ST-4/Chainpollion will descend to the comet's surface and will anchor itself with a 3-meter long harpoon. Scientific experiments include narrow and wide angle cameras for orbital mapping, panoramic and near-field cameras for landing site mapping, a gas chromatograph/mass spectrometer, a combined microscope and infrared spectrometer, and physical properties probes. Cometary samples will be obtained from depths up to 1.4 meters. The spacecraft is solar powered with rechargeable batteries, thus allowing a long duration mission on the nucleus surface. At the time of this writing, the ST-4/Champollion spacecraft was undergoing a major redesign to fit within NASA cost constraints, and approval of the mission is pending.

Weissman, Paul R.↗
DOE OSTI2020

Chemical trends of deep levels in van der Waals semiconductors

Abstract Properties of semiconductors are largely defined by crystal imperfections including native defects. Van der Waals (vdW) semiconductors, a newly emerged class of materials, are no exception: defects exist even in the purest materials and strongly affect their electrical, optical, magnetic, catalytic and sensing properties. However, unlike conventional semiconductors where energy levels of defects are well documented, they are experimentally unknown in even the best studied vdW semiconductors, impeding the understanding and utilization of these materials. Here, we directly evaluate deep levels and their chemical trends in the bandgap of MoS 2 , WS 2 and their alloys by transient spectroscopic study. One of the deep levels is found to follow the conduction band minimum of each host, attributed to the native sulfur vacancy. A switchable, DX center - like deep level has also been identified, whose energy lines up instead on a fixed level across different hosts, explaining a persistent photoconductivity above 400 K.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2021

Graphene U-Net v1

Graphene U-Net is a library that provides a simplified platform for training deep neural networks for the task of microscopy image segmentation. It contains functions and classes that make the process of training a neural network simple such that non-ML experts can train and evaluate models on their own datasets. It uses Pytorch as a backend and can run on both CUDA-enable GPUs and CPUs. It contains the main library file Microscopy_Unet.py as well as unet.py which contains the UNET model used for this software. This can be replaced with any other fully convolutional deep learning architecture with relative ease.

Sadre, Robbie↗
NASA NTRS1979

Deep radio occultations and 'evolute flashes' - Their characteristics and utility for planetary studies

Deep radio occultation signals from spacecraft passing behind planets may provide data on atmospheric absorption, turbulence, and structure, as well as information on the effects of planetary gravitational moments, rotation and zonal winds on the atmospheric shape. The strength of radio signals from a spacecraft passing behind a planet will at first decrease because of defocusing in the atmosphere, but then increase as the evolute of the planetary limb is neared, due to focusing caused by limb curvature within the evolute. Within the evolute, the availability of four simultaneous signal paths over four limb positions may render focused signals instantaneously great. The passage of Voyager 1 behind Jupiter and Voyager 2 behind Saturn will provide a test of deep radio occultation studies.

Eshleman, V. R.↗
DOE OSTI2020

Deep Generative Models that Solve PDEs: Distributed Computing for Training Large Data-Free Models

Recent progress in scientific machine learning (SciML) has opened up the possibility of training novel neural network architectures that solve complex partial differential equations (PDEs). Several (nearly data free) approaches have been recently reported that successfully solve PDEs, with examples including deep feed forward networks, generative networks, and deep encoder-decoder networks. However, practical adoption of these approaches is limited by the difficulty in training these models, especially to make predictions at large output resolutions (≥1024×1024). Here we report on a software framework for data parallel distributed deep learning that resolves the twin challenges of training these large SciML models - training in reasonable time as well as distributing the storage requirements. Our framework provides several out of the box functionality including (a) loss integrity independent of number of processes, (b) synchronized batch normalization, and (c) distributed higher-order optimization methods. We show excellent scalability of this framework on both cloud as well as HPC clusters, and report on the interplay between bandwidth, network topology and bare metal vs cloud. We deploy this approach to train generative models of sizes hitherto not possible, showing that neural PDE solvers can be viably trained for practical applications. We also demonstrate that distributed higher-order optimization methods are 2-3× faster than stochastic gradient-based methods and provide minimal convergence drift with higher batch-size.

PDEs↗
DOE OSTI2022

Differentiable, Learnable, Regionalized Process-Based Models With Multiphysical Outputs can Approach State-Of-The-Art Hydrologic Prediction Accuracy

Predictions of hydrologic variables across the entire water cycle have significant value for water resources management as well as downstream applications such as ecosystem and water quality modeling. Recently, purely data-driven deep learning models like long short-term memory (LSTM) showed seemingly insurmountable performance in modeling rainfall runoff and other geoscientific variables, yet they cannot predict untrained physical variables and remain challenging to interpret. Here, we show that differentiable, learnable, process-based models (called δ models here) can approach the performance level of LSTM for the intensively observed variable (streamflow) with regionalized parameterization. We use a simple hydrologic model HBV as the backbone and use embedded neural networks, which can only be trained in a differentiable programming framework, to parameterize, enhance, or replace the process-based model's modules. Without using an ensemble or post-processor, δ models can obtain a median Nash-Sutcliffe efficiency of 0.732 for 671 basins across the USA for the Daymet forcing data set, compared to 0.748 from a state-of-the-art LSTM model with the same setup. For another forcing data set, the difference is even smaller: 0.715 versus 0.722. Meanwhile, the resulting learnable process-based models can output a full set of untrained variables, for example, soil and groundwater storage, snowpack, evapotranspiration, and baseflow, and can later be constrained by their observations. Both simulated evapotranspiration and fraction of discharge from baseflow agreed decently with alternative estimates. The general framework can work with models with various process complexity and opens up the path for learning physics from big data.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2021

AI-Improved Resolution Projections of Population Characteristics and Imperviousness Can Improve Resolution and Accuracy of Urban Flood Predictions

Focal Area: (2) Predictive modeling through the use of AI techniques and AI-derived model components. Specifically, we call for the use of deep generative models from ML for providing high-resolution projections of impervious surface area, as well as neural network solvers for fast approximation of urban hydrodynamics to provide greatly enhanced forecasts of future urban flood dynamics.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2020

Event Selection in the MicroBooNE Deep Learning Based Low Energy Excess Analysis Using Two-Body Scattering Criteria

The uniquely detailed neutrino event information from liquid argon time projection chambers allows reconstruction of a set of kinematic quantities that over-constrain the expectations for charged current quasielastic scattering (CCQE). MicroBooNE makes use of the CCQE consistency requirements in a deep-learning-based search for the MiniBooNE low energy excess analysis. This requirement rejects backgrounds as well as events with poorly reconstructed neutrino energy due to final state interactions of the outgoing proton. The results presented here demonstrate the quality of the selection of $\nu_e$ and $\nu_µ$ events. We show excellent agreement between the data and the simulation across many data sets. This positions us to be ready to unblind the MicroBooNE low energy excess analysis in the very near future.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
NASA NTRS1980

The double quasar Q0957 + 561 A, B - A gravitational lens image formed by a galaxy at z = 0.39

Deep CCD pictures show a rich cluster of galaxies around the double quasar Q0957 + 561 as well as faint extended wings on the image of the southern quasar. Calculations of gravitational imaging by King model mass distributions show that the cluster and the brightest galaxy together, acting as a gravitational lens on the light from a single, more distant QSO, can easily reproduce all of the present observations. It is concluded that the double quasar is almost certainly the multiple image of a single object produced by a gravitational lens.

Young, P.↗
NASA NTRS1984

Atomic inner-shell transitions

Atomic inner-shell processes have quite different characteristics, in several important aspects, from processes in the optical regime. Energies are large, e.g., the 1s binding energy reaches 100 keV at Z = 87; relativistic and quantum-electrodynamic effects therefore are strong. Radiationless transitions vastly dominate over photon emission in most cases. Isolated inner-shell vacancies have pronounced single-particle character, with correlations generally contributing only approximately 1 eV to the 1s and 2p binding energies; the structure of such systems is thus well tractable by independent-particle self-consistent-field atomic models. For systems containing multiple deep inner-shell vacancies, or for highly stripped ions, the importance of relativistic intermediate coupling and configuration interaction becomes pronounced. Cancellation of the Coulomb interaction can lead to strong manifestations of the Breit interaction in such phenomena as multiplet splitting and hypersatellite X-ray shifts. Unique opportunities arise for the test of theory.

Crasemann, B.↗
NASA NTRS1987

A swept wing panel in a low speed flexible walled test section

The testing of two-dimensional airfoil sections in adaptive wall tunnels is relatively widespread and has become routine at all speeds up to transonic. In contrast, the experience with the three-dimensional testing of swept panels in adaptive wall test sections is very limited, except for some activity in the 1940's at NPL, London. The current interest in testing swept wing panels led to the work covered by this report, which describes the design of an adaptive-wall swept-wing test section for a low speed wind tunnel and gives test results for a wing panel swept at 40 deg. The test section has rigid flat sidewalls supporting the panel, and features flexible top and bottom wall with ribs swept at the same angle as the wing. When streamlined, the walls form waves swept at the same angle as the wing. The C sub L (-) curve for the swept wing, determined from its pressure distributions taken with the walls streamlined, compare well with reference data which was taken on the same model, unswept, in a test section deep enough to avoid wall interference.

Goodyer, M. J.↗
NASA NTRS1990

The atmospheric structure and dynamical properties of Neptune derived from ground-based and IUE spectrophotometry

A wide range of recent full-disk spectral observations is used to constrain the atmospheric structure and dynamical properties of Neptune; analytical determinations are made of the abundances of such spectrally active gas species as the deep-atmosphere CH4 molar fraction and the mean ortho/para hydrogen ratio in the visible atmosphere, as well as stratospheric and tropospheric aerosol properties. Compared to Uranus, the greater abundance and shorter lifetimes of Neptunian particulates in the stratospheric region irradiated by the solar UV flux indicate that such radiation is the darkening agent of stratospheric aerosols on both planets.

Baines, Kevin H.↗
NASA NTRS1991

Mean flow and variability in the Kuroshio Extension from Geosat altimetry data

The mean flow and temporal and spatial variations of the Kuroshio Extension in the region of 140-180 deg E and 30-40 deg N are investigated using altimeter data from the Geosat Exact Repeat Mission (ERM). Mean surface height profiles are estimated along individual tracks by assuming the velocity profile of the Kuroshio Extension to be Gaussian-shaped and by successively fitting this synthetic current's height profile to the residual height data. The mean profiles from ascending and descending tracks are used to derive the mean surface height by an inverse method and to obtain the absolute surface height fields for the first 2.5 yr of the Geosat ERM. Both the mean and the instantaneous height fields thus derived compare well with the available hydrographic data and the SST patterns from the NOAA satellites. Effects of deep mean flow and baroclinic shear are found to be important in explaining the observed propagation speeds.

Qiu, BO↗
NASA NTRS2001

Global Bathymetric Prediction For Ocean Modeling and Marine Geophysics

We proposed to construct a complete bathymetric map of the oceans at a 3-10 km resolution by combining all of the available depth soundings collected over the past 30 years with high resolution marine gravity information provided by the Geosat, ERS-1/2, and Topex/Poseidon altimeters. Detailed bathymetry is essential for understanding physical oceanography and marine geophysics. Currents and tides are controlled by the overall shapes of the ocean basins as well as the smaller sharp ocean ridges and seamounts. Because erosion rates are low in the deep oceans, detailed bathymetry reveals the mantle convection patterns, the plate boundaries, the cooling/subsidence of the oceanic lithosphere, the oceanic plateaus, and the distribution of off-ridge volcanoes. We proposed to: (1) Accumulate all available depth soundings collected over the past 30 years; (2) Use the short wavelength (< 160 km) satellite gravity information to interpolate between sparse ship soundings; (3) Improve the resolution of the marine gravity field using enhanced estimates along repeat altimeter profiles together with the dense altimeter measurements; (4) Refine/improve bathymetric predictions using the improved resolution gravity field and also by investigating computer-intensive methods for bathymetric prediction such as inverse theory; and (5) Produce a 'Globe of the Earth' similar to the globe of Venus prepared by the NASA Magellan investigation. This will also include the best available digital land data.

Sandwell, David T.↗
NASA NTRS2001

Carbon Nanotube Scanning Probe for Surface Profiling of DUV and 193 nm Photoresist Pattern

The continual scaling down of semiconductors to 100 nm and below necessitates a characterization technique to resolve high aspect ratio features in the nanoscale regime. This paper reports the use of atomic force microscope coupled with high aspect ratio multi-walled carbon nanotube scanning probe tip for the purpose of imaging surface profile of photoresists. Multi-walled carbon nanotube tips used in this work are 5-10 nm in diameter and about a micron long. Their exceptional mechanical strength and ability to reversibly buckle enable to resolve steep, deep nanometer-scale features. Images of photoresist patterns generated by 257 nm interference lithography as well as 193 nm lithography are presented to demonstrate multi-walled carbon nanotube scanning probe tip for applications in metrology.

Nguyen, Cattien V.↗