Search NASA⌕ Search

SEARCH · Search NASA

Results for “Deep Material Network”

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 127 records · Page 7
All sourcesNASA NTRSDOE OSTI
NASA NTRS2013

High-Reliability Waveguide Vacuum/Pressure Window

The NASA Deep Space Network (DSN) uses commercial waveguide windows on the output waveguide of Ka-band (32 GHz) low-noise amplifiers. Mechanical failure of these windows resulted in an unacceptable loss in tracking time. To address this issue, a new Ka-band WR-28 waveguide window has been designed, fabricated, and tested. The window uses a slab of low-loss, low-dielectric constant foam that is bonded into a 1/2-wave-thick waveguide/flange. The foam is a commercially available, rigid, closed-cell polymethacrylimide. It has excellent electrical properties with a dielectric constant of 1.04, and a loss tangent of 0.01. It is relatively strong with a tensile strength of 1 MPa. The material is virtually impermeable to helium. The finished window exhibits a leak rate of less than 3x10(exp -3)cu cm/s with helium. The material is also chemically resistant and can be cleaned with acetone. The window is constructed by fabricating a window body by brazing a short length of WR-28 copper waveguide into a standard rectangular flange, and machining the resulting part to a thickness of 4.6 mm. The foam is machined to a rectangular shape with a dimension of 7.06x3.53 mm. The foam is bonded into the body with a two-part epoxy. After curing, the excess glue and foam are knife-trimmed by hand. The finished window has a loss of less than 0.08 dB (2%) and a return loss of greater than 25 dB at 32 GHz. This meets the requirements for the DSN application. The window is usable for most applications over the entire 26-to-40-GHz waveguide band. The window return loss can be tuned to a required frequency by var y in g the thickness of the window slightly. Most standard waveguide windows use a thin membrane of material bonded into a recess in a waveguide flange, or sandwiched between two flanges with a polymer seal. Designs using the recessed window are prone to mechanical failure over time due to constraints on the dimensions of the recess that allow the bond to fail. Designs using the sandwich method are often permeable to helium, which prohibits the use of helium leak detection. At the time of this reporting, 40 windows have been produced. Twelve are in operation with a combined operating time of over 30,000 hours without a failure.

Britcliffe, Michael J.↗
DOE OSTI2026

A Statistician’s Overview of Physics-Informed Neural Networks for Spatio-Temporal Data

The recent success of deep neural network models with physical constraints (so-called, Physics-Informed Neural Networks, PINNs) has led to renewed interest in the incorporation of mechanistic information in predictive models. Statisticians and others have long been interested in this problem, which has led to several practical and innovative solutions dating back decades. In this overview, we focus on the problem of data-driven prediction and inference of dynamic spatio-temporal processes that include mechanistic information, such as would be available from partial differential equations, with a strong focus on the quantification of uncertainty associated with data, process, and parameters. Here, we give a brief review of several paradigms and focus our attention on Bayesian implementations given they naturally accommodate uncertainty quantification. We then show that it is straight-forward to include the Bayesian PINN (B-PINN) within the Bayesian hierarchical model (BHM) framework that has long been considered for modeling dynamic spatio-temporal processes. Such a BHM-PINN is illustrated via a simulation study in which a latent nonlinear Burgers’ equation PDE governs the dynamics of Poisson distributed spatio-temporal data. Supplementary materials for this article are available online, including a standardized description of the materials available for reproducing the work.

Bayesian↗
DOE OSTI2026

Detection and imaging of chemicals and hidden explosives using terahertz time-domain spectroscopy and deep learning

Detecting concealed chemicals and explosives remains a critical challenge in global security. Terahertz time-domain spectroscopy (THz-TDS) offers a promising non-invasive and stand-off detection technique owing to its ability to penetrate optically opaque materials without causing ionization damage. While many chemicals exhibit distinct spectral features in the terahertz range, conventional terahertz-based detection methods often struggle in real-world environments, where variations in sample geometry, thickness, and packaging can lead to inconsistent spectral responses. In this study, we present a chemical imaging system that integrates THz-TDS with deep learning to enable accurate pixel-level identification and classification of different explosives. Operating in reflection mode and enhanced with plasmonic nanoantenna arrays, our THz-TDS system achieves a peak dynamic range of 96 dB and a detection bandwidth of 4.5 THz, supporting practical, stand-off operation. By analyzing individual time-domain pulses with deep neural networks, the system exhibits strong resilience to environmental variations and sample inconsistencies. Blind testing across eight chemicals—including pharmaceutical excipients and explosive compounds—resulted in an average classification accuracy of 99.42% at the pixel level. Notably, the system maintained an average accuracy of 88.83% when detecting explosives concealed under opaque paper coverings, demonstrating its robust generalization capability. These results highlight the potential of combining advanced terahertz spectroscopy with neural networks for highly sensitive and specific chemical and explosive detection in diverse and operationally relevant scenarios.

Imaging and sensing↗
DOE OSTI2024

Using the Principals of Electrochemistry to Understand and Overcome the Complicated Degradation Mechanisms of Silicon Anodes in Lithium-Ion Batteries [Slides]

Battery technology is the most significant problem facing widespread market adoption of battery electric vehicles (BEVs). The low energy density of state-of-the-art lithium-ion electrode materials leaves BEV owners and prospective buyers with lower ranges than a comparable internal combustion engine (ICE) vehicle, and vulnerable to a nascent fast charging network. Silicon has the potential to increase the anode energy density by nearly ten times compared to the incumbent material (graphite) and make BEVs a more competitive transportation option. However, lithiated silicon is extremely reactive towards components of the electrolyte and forms a heterogeneous, complicated, and dynamic solid at its surface known as the solid electrolyte interphase (SEI). Ideally, the SEI would passivate the silicon surface, but continuous chemical degradation persists even when the battery is not operating. This reactivity reduces silicon anode lifetimes well below the necessary standards for BEVs. The NREL-led Silicon Consortium Project is dedicated to understanding and solving these mechanisms of degradation. Here, I will discuss an electrochemical method that provides deep insights into the silicon interface during battery operation. I will link these observations to fundamental electrochemical principals and how they translate into actionable strategies that extend the lifetime of silicon anodes.

25 ENERGY STORAGE↗
DOE OSTI2025

Simultaneous prediction of structural properties in epitaxially–grown GaN with quantum and conventional multi–output learning algorithms

Hundreds of GaN thin film crystal plasma–assisted molecular beam epitaxy synthesis experiment records spanning two decades were organized into a dataset correlating the growth experiment design parameters with discrete, binary determinations of crystallinity and surface morphology. Conventional data science techniques as well as both quantum and classical multi–output supervised machine learning algorithms were implemented to investigate the relationships between the operating parameter data and the structural figures of merit. Correlation coefficients, decision tree nodes, p–values, and SHAP values all support substrate temperature and gallium effusion cell conditions as being statistically significant for simultaneously influencing GaN crystallinity and surface morphology. Here, a conventional deep neural network learned best from the data, followed by a quantum–classical hybrid gradient boosting algorithm. When combined with calculations of uncertainty intervals based on VennAbers predictors, machine learning predictions of both structural properties show good agreement with results reported in published experimental literature.

36 MATERIALS SCIENCE↗
DOE OSTI2024

Deep-learning based artificial intelligence tool for melt pools and defect segmentation

Accelerating fabrication of additively manufactured components with precise microstructures is important for quality and qualification of built parts, as well as for a fundamental understanding of process improvement. Accomplishing this requires fast and robust characterization of melt pool geometries and structural defects in images. This paper proposes a pragmatic approach based on implementation of deep learning models and self-consistent workflow that enable systematic segmentation of defects and melt pools in optical images. Deep learning is based on an image-to-image translation–conditional generative adversarial neural network architecture. An artificial intelligence (AI) tool based on this deep learning model enables fast and incrementally more accurate predictions of the prevalent geometric features, including melt pool boundaries and printing-induced structural defects. We present statistical analysis of geometric features that is enabled by the AI tool, showing strong spatial correlation of defects and the melt pool boundaries. The correlations of widths and heights of melt pools with dataset processing parameters show the highest sensitivity to thermal influences resulting from laser passes in adjacent and subsequent layer passes. The presented models and tools are demonstrated on the aluminum alloy and datasets produced with different sets of processing parameters. However, they have universal quality and could easily be adapted to different material compositions. The method can be easily generalized to microstructural characterizations other than optical microscopy.

additive manufacturing↗
DOE OSTI2025

Accelerating charge estimation in molecular dynamics simulations using physics-informed neural networks: corrosion applications

Molecular Dynamics (MD) simulations are used to understand the effects of corrosion on metallic materials in salt brine. Reactive force fields in classical MD enable accurate modeling of bond formation and breakage in the aqueous medium and at the metal-electrolyte interface, while also facilitating dynamic partial charge equilibration. However, MD simulations are computationally intensive and unsuitable for modeling the long time scales characteristic of corrosive phenomena. To address this, we develop reduced-order machine learning models that provide accurate and efficient predictions of charge density in corrosive environments. Specifically, we use Long Short-Term Memory (LSTM) networks to forecast charge density evolution based on atomic environments represented by Smooth Overlap of Atomic Positions (SOAP) descriptors. A physics-informed loss function enforces charge neutrality and electronegativity equivalence. The atomic charges predicted by the deep learning model trained on this work were obtained two orders of magnitude faster than those from molecular dynamics (MD) simulations, with an error of less than 3% compared to the MD-obtained charges, even in extrapolative scenarios, while adhering to physical constraints. This demonstrates the excellent accuracy, computational efficiency, and validity of the developed model. Lastly, even though developed for corrosion, these protocols are formulated in a phenomenon-agnostic manner, allowing application to various variable-charge interatomic potentials and related fields.

Atomistic models↗
NASA NTRS2010

Jet Propulsion Laboratory: Annual Report 2009

2009 was truly the year of astronomy at the Jet Propulsion Laboratory. While the world at large was celebrating the International Year of Astronomy, we were sending more telescopes into space than in any other year, ever. As these missions unfold, the astronomers are sure to change the way we see the universe. One of the newly lofted observatories is on a quest to find planets like our own Earth orbiting other stars. Another is a telescope that gathers infrared light to help discover objects ranging from near-Earth asteroids to galaxies in the deepest universe. We also contributed critical enabling technologies to yet two other telescopes sent into space by our partners in Europe. And astronauts returned to Earth with a JPL-built camera that had captured the Hubble Space Telescope's most memorable pictures over many years. And while it was an epic time for these missions, we were no less busy in our other research specialties. Earth's moon drew much attention from our scientists and engineers, with two JPL instruments riding on lunar orbiters; previously unseen views of shadowed craters were provided by radar imaging conducted with the giant dish antennas of the Deep Space Network, our worldwide communication portal to spacecraft around the solar system. At Mars, our rovers and orbiters were highly productive, as were missions targeting Saturn, comets and the asteroid belt. Here at our home planet, satellites and instruments continued to serve up important information on global climate change. But our main business is, of course, exploring. Many initiatives will keep us busy for years. In 2009, NASA gave approval to start planning a major flagship mission to Jupiter's moon Europa in search of conditions that could host life, working with our partners in Europe. In addition to our prospective Earth science projects, we have full slates of missions in Mars exploration, planetary exploration and space-based astronomy. This year's annual report continues our recent direction of recounting the Laboratory's accomplishments throughout the year month by month.

Jet Propulsion Laboratory (JPL)↗
DOE OSTI2024

Deep learning with mixup augmentation for improved pore detection during additive manufacturing

In additive manufacturing (AM), process defects such as keyhole pores are difficult to anticipate, affecting the quality and integrity of the AM-produced materials. Hence, considerable efforts have aimed to predict these process defects by training machine learning (ML) models using passive measurements such as acoustic emissions. This work considered a dataset in which keyhole pores of a laser powder bed fusion (LPBF) experiment were identified using X-ray radiography and then registered both in space and time to acoustic measurements recorded during the LPBF experiment. Due to AM’s intrinsic process controls, where a pore-forming event is relatively rare, the acoustic datasets collected during monitoring include more non-pores than pores. In other words, the dataset for ML model development is imbalanced. Moreover, this imbalanced and sparse data phenomenon remains ubiquitous across many AM monitoring schemes since training data is nontrivial to collect. Hence, we propose a machine learning approach to improve this dataset imbalance and enhance the prediction accuracy of pore-labeled data. Specifically, we investigate how data augmentation helps predict pores and non-pores better. This imbalance is improved using recent advances in data augmentation called Mixup, a weak-supervised learning method. Convolutional neural networks (CNNs) are trained on original and augmented datasets, and an appreciable increase in performance is reported when testing on five different experimental trials. When ML models are trained on original and augmented datasets, they achieve an accuracy of 95% and 99% on test datasets, respectively. We also provide information on how dataset size affects model performance. Lastly, we investigate the optimal Mixup parameters for augmentation in the context of CNN performance.

42 ENGINEERING↗
NASA NTRS2007

Jet Propulsion Laboratory: Annual Report 2006

Nothing is as gratifying in space exploration as when we are surprised by the unexpected. Much of our work progresses in an orderly way, from concept to plan to creation to finding. But now and then we are caught off-guard by something startlingly new, and it is these moments that make our hearts race and leave us with many of our most compelling memories. And 2006 was an exceptional year for the unforeseen. One of our orbiters shocked many with stark proof that liquid water, the seemingly long-gone force that reshaped so much of the scenery of Mars, still flows there today,at least in occasional bursts. Another spacecraft caught us by surprise with photos of Yellowstone-like geysers on one of Saturn's seemingly nondescript moons, Enceladus. A spaceborne observatory created to plumb the life histories of stars and galaxies showed off a completely unexpected talent when it revealed the day and night faces of a fire and ice planet far beyond our solar system 40 light-years away. A newly launched Earth observer revealed that the clouds that decorate our own planet are not what we thought them to be in many ways. Of course, not all of the high points of the year arrived on our doorstep in such unexpected ways. There was also great drama when missions came off exactly as planned, such as when Stardust's sample return capsule made a flawless landing in the Utah desert, bringing home samples of cometary and interstellar dust. Mars Reconnaissance Orbiter slipped into orbit around the red planet exactly as planned. Numerous other missions and technology programs likewise made great achievements during the year. In all, 17 spacecraft and six instruments were stationed across the solar system, studying our own world, other planets, comets and the deeper universe. All of these achievements were enabled by many teams and systems at the Laboratory. The Deep Space Network of communications complexes across three continents supported all of NASA's solar system missions, and several from our international partners, while looking ahead to reinvent itself for the future. Technologists were at work creating innovations both for NASA missions and terrestrial uses. JPL's community of scientific researchers was equally busy coordinating the science activities of our missions and pursuing independent investigations. None of this would be possible without the support of business and administrative teams dedicated to making the Laboratory's institutional environment as world-class as its technical face, or the public engagement specialists who bring the experience of space to the country's diverse publics. All of our missions in one way or another support our nation's Vision for Space Exploration, which envisages a gradually widening robotic and human presence across the solar system in the years ahead.

National Aeronautics and Space Administration (NAS↗
NASA NTRS2013

Pyroclastic Deposits in the Floor-fractured Crater Alphonsus

Alphonsus, the 118 km diameter floor-fractured crater, is located immediately east of Mare Nubium. Eleven pyroclastic deposits have been identified on the crater's floor. Early telescopic spectra suggest that the floor of Alphonsus is noritic, and that the pyroclastic deposits contain mixtures of floor material and a juvenile component including basaltic glass. Head and Wilson contend that Nubium lavas intruded the breccia zone beneath Alphonsus, forming dikes and fractures on the crater floor. In this model, the magma ascended to the level of the mare but cooled underground, and a portion broke thru to the surface in vulcanian (explosive) eruptions. Alternatively, the erupted material could be from a source unrelated to the mare, in the style of regional pyroclastic deposits. High-resolution images and spectroscopy from the Moon Mineralogy Mapper (M3), Diviner Lunar Radiometer, and Lunar Reconnaissance Orbiter Camera Narrow Angle Camera (NAC) provide data to test these formation models. Spectra from M3 confirm that the crater floor is primarily composed of noritic material, and that the Nubium lavas are basaltic. Spectra from the three largest pyroclastic deposits in Alphonsus are consistent with a minor low- Ca pyroxene component in a glass-rich matrix. The centers of the 2 micron absorption bands have wavelengths too short to be of the same origin as the Nubium basalts. Diviner Christiansen feature (CF) values were used to estimate FeO abundances for the crater floor, Nubium soil, and pyroclastic deposits. The estimated abundance for the crater floor (7.5 +/- 1.4 wt.%) is within the range of FeO values for Apollo norite samples. However, the estimated FeO abundance for Nubium soil (13.4 +/- 1.4 wt.%) is lower than those measured in most mare samples. The difference may reflect contamination of the mare soil by highland ejecta. The Diviner-derived FeO abundance for the western pyroclastic deposit is 13.8 +/- 3.3 wt.%. This is lower than the values for mare soil samples, but within the range of analyzed pyroclastic glasses. The NAC images of the pyroclastic vents highlight their bright wall materials. The M3 spectra of the southeastern vent indicate that this bright material is noritic, likely crater floor material exposed by explosive eruption. These observations address the hypothesis that Nubium lavas intruded the fracture network beneath Alphonsus, leading to localized vulcanian-style eruptions. This model implies that the eruption products should be dominated by crystalline basalt fragments similar in elemental composition and mineralogy to mare lavas. The bright noritic material exposed in the vent walls is consistent with explosive eruptions. The estimated FeO abundances for the pyroclastic deposits are too low to be consistent with FeO abundances measured in mare basalts, but are within the range of pyroclastic glass samples. The visible- to near-infrared (VIS-NIR) spectra of the pyroclastic deposits and Nubium soils are significantly different, suggesting that the pyroclastics are unrelated to the mare basalts. The pyroclastic spectra are consistent with Fe-bearing glass plus small amounts of noritic wall rock. Similar glassy materials dominate regional pyroclastic deposits, suggesting a deep source for the pyroclastics observed in Alphonsus.

Allen, Carlton C.↗
DOE OSTI2025

A self-supervised robotic system for autonomous contact-based spatial mapping of semiconductor properties

Integrating robotically driven contact-based material characterization techniques into self-driving laboratories can enhance measurement quality, reliability, and throughput. While deep learning models support robust autonomy, current methods lack reliable pixel-precision positioning and require extensive labeled data. To overcome these challenges, we propose an approach for building self-supervised autonomy into contact-based robotic systems that teach the robot to follow domain expert measurement principles at high throughputs. We demonstrate the performance of this approach by autonomously driving a 4-DOF robotic probe for 24 hours to characterize semiconductor photoconductivity at 3025 uniquely predicted poses across a gradient of drop-casted perovskite film compositions, achieving throughputs of more than 125 measurements per hour. Spatially mapping photoconductivity onto each drop-casted film reveals compositional trends and regions of inhomogeneity, valuable for identifying manufacturing defects. With this self-supervised neural network–driven robotic system, we enable high-precision and reliable automation of contact-based characterization techniques at high throughputs, thereby allowing measurement of previously inaccessible yet important semiconductor properties for self-driving laboratories.

Science & Technology - Other Topics↗
NASA NTRS2013

NASA Tech Briefs, Februrary 2013

Topics covered include: Measurements of Ultra-Stable Oscillator (USO) Allan Deviations in Space; Gaseous Nitrogen Orifice Mass Flow Calculator; Validation of Proposed Metrics for Two-Body Abrasion Scratch Test Analysis Standards; Rover Low Gain Antenna Qualification for Deep Space Thermal Environments; Automated, Ultra-Sterile Solid Sample Handling and Analysis on a Chip; Measuring and Estimating Normalized Contrast in Infrared Flash Thermography; Spectrally and Radiometrically Stable, Wideband, Onboard Calibration Source; High-Reliability Waveguide Vacuum/Pressure Window; Methods of Fabricating Scintillators With Radioisotopes for Beta Battery Applications; Magnetic Shield for Adiabatic Demagnetization Refrigerators (ADR); CMOS-Compatible SOI MESFETS for Radiation-Hardened DC-to-DC Converters; Silicon Heat Pipe Array; Adaptive Phase Delay Generator; High-Temperature, Lightweight, Self-Healing Ceramic Composites for Aircraft Engine Applications; Treatment to Control Adhesion of Silicone-Based Elastomers; High-Temperature Adhesives for Thermally Stable Aero-Assist Technologies; Rockballer Sample Acquisition Tool; Rock Gripper for Sampling, Mobility, Anchoring, and Manipulation; Advanced Magnetic Materials Methods and Numerical Models for Fluidization in Microgravity and Hypogravity; Data Transfer for Multiple Sensor Networks Over a Broad Temperature Range; Using Combustion Synthesis to Reinforce Berms and Other Regolith Structures; Visible-Infrared Hyperspectral Image Projector; Three-Axis Attitude Estimation With a High-Bandwidth Angular Rate Sensor Change_Detection.m; AGATE: Adversarial Game Analysis for Tactical Evaluation; Ionospheric Simulation System for Satellite Observations and Global Assimilative; Modeling Experiments (ISOGAME); An Extensible, User- Modifiable Framework for Planning Activities; Mission Operations Center (MOC) - Precipitation Processing System (PPS) Interface Software System (MPISS); Automated 3D Damaged Cavity Model Builder for Lower Surface Acreage Tile on Orbiter; Mixed Linear/Square-Root Encoded Single-Slope Ramp Provides Low-Noise ADC with High Linearity for Focal Plane Arrays; RUSHMAPS: Real-Time Uploadable Spherical Harmonic Moment Analysis for Particle Spectrometers; Powered Descent Guidance with General Thrust-Pointing Constraints; X-Ray Detection and Processing Models for Spacecraft Navigation and Timing; and Extreme Ionizing-Radiation-Resistant Bacterium

Source record↗
DOE OSTI2025

Constrained GAN-Generated X-Ray CT Data For Self-Supervised And Foundation-Model Segmentation Of Concrete Microstructures

Three-dimensional characterization of materials using X-ray computed tomography (XCT) is challenging due to the complexity of internal structures, noise, and variations in resolution. Traditional computer vision models often struggle to accurately segment these images, particularly in domain-specific applications like materials science. While supervised deep learning approaches have been developed to address the limitations of conventional algorithms, they typically require large amounts of labeled training data and often fail to generalize across different datasets. Self-supervised, few-and zero-shot learning methods have gained prominence in natural image processing and segmentation tasks, but their application to scientific imaging remains limited due to the unique structural complexity, noise, and textural artifacts present in materials science data. In this work, we investigate how domain adaptation, leveraging physics-based and GAN-generated synthetic data, impacts segmentation performance. We introduce a modified Contrastive Unpaired Translation (CUT) model designed to generate realistic labeled data, which can be used for training, pre-training, and fine-tuning segmentation models for real XCT microstructure data. We evaluate the performance of two segmentation approaches: a self-supervised network (SSL-ALPNet) and a foundation model (Segment Anything Model), assessing their improvements when pre-trained and/or fine-tuned on the synthesized data. Our results demonstrate that leveraging synthetic data significantly enhances segmentation performance, particularly in challenging materials science applications.

Ziabari, Amir [ORNL] (ORCID:000000034776457X)↗
NASA NTRS2018

CubeX: A Compact X-Ray Telescope Enables Both X-Ray Fluorescence Imaging Spectroscopy and Pulsar Timing Based Navigation

This paper describes the benefits of a miniaturized X-ray telescope payload in the context of a lunar mission. The first part describes the payload in detail, the second part summarizes a small satellite mission concept that utilizes its compact form factor and performance. The CubeX instrument can be used for both X-ray fluorescence (XRF) imaging spectroscopy and X-ray pulsar timing based navigation (XNAV). Using our recent technological advances in X-ray optics and sensors, CubeX combines high angular resolution (<1 arcminutes) Miniature Wolter-I X-ray optics (MiXO) with a common focal plane consisting of high spectral resolution (<150 eV at 1 keV) CMOS X-ray sensors and a high timing resolution (< 1 usec) SDD X-ray sensor. This novel combination of the instruments enables both XRF measurements and XNAV operations without moving parts. The high angular resolution of the MiXO opens a wide range of orbital configurations for observation. Given that performance, the instrument has unprecedented small volume (~1×1×6U), mass (<6 kg), and power (<9W) requirements and opens a wide range of applications for a variety of targets and missions including NEOs and Martian moons. In this paper we illustrate one potential application for a lunar mission concept: The elemental composition of the Moon holds keys to understanding the origin and evolution of both the Moon and the Earth. X-ray fluorescence (XRF), induced either by solar X-ray flux or energetic ions, carries decisive signatures of surface elemental composition. X-ray observations, therefore, give a unique, powerful diagnostic tool for remotely determining elemental abundances including major rock forming elements such as Mg, Al, Na, Si, Fe, and Ca. Through high-resolution XRF imaging spectroscopy, CubeX searches for small patches of elusive lower crust and mantle material excavated within and around impact craters. CubeX identifies regional compositional variations and allows straightforward comparison of elemental distributions with the surface topography from LRO and the gravity data from GRAIL. The elemental compositions of the lower crust and the mantle are sensitive to the conditions of the giant impact which led to the Moon's formation and the subsequent lunar magma ocean (LMO), and thus they are key missing pieces in understanding the formation and early evolution of the Moon. In between XRF observations, CubeX also leverages the technology of high resolution X-ray imaging and time series measurements to conduct XNAV operations and evaluate their performance. Deep space navigation is a critical issue for small planetary missions. XNAV can enable low-cost autonomous deep-space navigation, and has the potential to greatly assist, or even outperform, NASA's Deep Space Network (DSN) or ESA's European Space Tracking (ESTRACK). CubeX is designed to perform sequential observations of 3-4 millisecond pulsars (MSPs) to solve the spacecraft trajectory for absolute navigation, and explore the remote sensing capability of XNAV. In the presented mission concept, the Moon's relative proximity enables a straightforward evaluation of the XNAV performance through DSN.

x-ray spectroscopy↗
DOE OSTI2024

Highly sensitive 2D X-ray absorption spectroscopy via physics informed machine learning

Abstract Improving the spatial and spectral resolution of 2D X-ray near-edge absorption structure (XANES) has been a decade-long pursuit to probe local chemical reactions at the nanoscale. However, the poor signal-to-noise ratio in the measured images poses significant challenges in quantitative analysis, especially when the element of interest is at a low concentration. In this work, we developed a post-imaging processing method using deep neural network to reliably improve the signal-to-noise ratio in the XANES images. The proposed neural network model could be trained to adapt to new datasets by incorporating the physical features inherent in the latent space of the XANES images and self-supervised to detect new features in the images and achieve self-consistency. Two examples are presented in this work to illustrate the model’s robustness in determining the valence states of Ni and Co in the LiNi x Mn y Co 1-x-y O 2 systems with high confidence.

36 MATERIALS SCIENCE↗
NASA NTRS2006

Jet Propulsion Laboratory: Annual Report 2005

What an amazing host of new sights 2005 brought us. With impeccable choreography, one spacecraft sent an impactor slamming into a comet, reversing the traditional view of these space wayfarers by revealing it to be buried in deep drifts of a fine talcum-like powder. Another spacecraft delivered a probe from our European partners to the surface of Saturn's haze-shrouded moon Titan, disclosing a landscape eerily like Earth's, if we had methane rivers cascading down hillsides of ice. An orbiting observatory for the first time showed us the light from planets circling other stars, which astronomers previously knew to exist only from indirect clues. Throughout the year we also amassed continually expanding views of Earth as well as Mars, by far the most-explored planet after our own. In all, 18 spacecraft and five instruments were stationed across the solar system, studying our own world, other planets, comets and the deeper universe. These missions were enabled by the efforts of everyone at JPL. The Deep Space Network of communications complexes across three continents continued to experience a period of remarkable activity. Others were at work creating technologies both for NASA missions and other uses. JPL's contingent of scientific researchers was equally busy coordinating the science activities of our missions or pursuing independent investigations. None of this would be possible without the support of world-class business and administrative teams. All of our missions in one way or another support our nations Vision for Space Exploration, which envisages a gradually widening robotic and human presence across the solar system in the years ahead. The year was not without its challenges. NASA set forth to implement the Vision for Space Exploration, which resulted in some flight projects and technology efforts being terminated. To adjust to this new direction, it was necessary for us to reduce the JPL workforce by about five percent. Taking steps like this is painful, but we tried to make the process as orderly as possible. In the end, the adjustments made have left JPL on a healthy footing for the years ahead.

Jet Propulsion Laboratory (JPL)↗
DOE OSTI2024

A framework for strategic discovery of credible neural network surrogate models under uncertainty

The widespread integration of deep neural networks in developing data-driven surrogate models for high-fidelity simulations of complex physical systems highlights the critical necessity for robust uncertainty quantification techniques and credibility assessment methodologies, ensuring the reliable deployment of surrogate models in consequential decision-making. Here, this study presents the Occam Plausibility Algorithm for surrogate models (OPAL-surrogate), providing a systematic framework to uncover predictive neural network-based surrogate models within the large space of potential models, including various neural network classes and choices of architecture and hyperparameters. The framework is grounded in hierarchical Bayesian inferences and employs model validation tests to evaluate the credibility and prediction reliability of the surrogate models under uncertainty. Leveraging these principles, OPAL-surrogate introduces a systematic and efficient strategy for balancing the trade-off between model complexity, accuracy, and prediction uncertainty. The effectiveness of OPAL-surrogate is demonstrated through two modeling problems, including the deformation of porous materials for building insulation and turbulent combustion flow for ablation of solid fuels within hybrid rocket motors.

42 ENGINEERING↗