SEARCH · Search NASA
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.
Improve the quantitative results obtained from 3D gamma-ray imagers for use in safeguards inspections. In FY23 the work focused on continuing to improve the numerical results obtained with gamma-ray imagers and explore approaches to project such data onto 3D structures mapped with Self Localization and Mapping (SLAM) hardware and software.
We use machine-learning techniques to classify galaxy merger stages, which can unveil physical processes that drive the star formation and active galactic nucleus (AGN) activities during galaxy interaction. The sample contains 4690 galaxies from the integral field spectroscopy survey SDSS-IV MaNGA and can be separated into 1060 merging galaxies and 3630 nonmerging or unclassified galaxies. For the merger sample, there are 468, 125, 293, and 174 galaxies (1) in the incoming pair phase, (2) in the first pericentric passage phase, (3) approaching or just passing the apocenter, and (4) in the final coalescence phase or post-mergers. With the information of projected separation, line-of-sight velocity difference, Sloan Digital Sky Survey (SDSS) gri images, and MaNGA Hα velocity map, we are able to classify the mergers and their stages with good precision, which is the most important score to identify interacting galaxies. For the two-phase classification (binary; nonmerger and merger), the performance can be high (precision > 0.90) with LGBMClassifier . We find that sample size can be increased by rotation, so the five-phase classification (nonmerger, and merger stages 1, 2, 3, and 4) can also be good (precision > 0.85). The most important features come from SDSS gri images. The contribution from the MaNGA Hα velocity map, projected separation, and line-of-sight velocity difference can further improve the performance by 0%–20%. In other words, the image and the velocity information are sufficient to capture important features of galaxy interactions, and our results can apply to all the MaNGA data, as well as future all-sky surveys.
This exploratory project, entitled “Non-Destructive In-process Assessment of Thermal Spray Repairs,” evaluated the capabilities and limitations of four non-destructive testing (NDE) technologies for detection of defects in as-sprayed thermal spray coatings: acousto-ultrasonics, vibro-thermography, thermal wave infrared (IR) imaging, and flash IR thermography. The project goal was to identify a single technology suitable for use as an in-process inspection for two common classes of defects in thermal spray coatings, porosity and disbonds. Two rounds of test coupons were fabricated with intentionally seeded defects in AISI 420 stainless steel coatings deposited by twin wire arc (TWA) spray onto gray cast iron substrates. These standardized test coupons allowed the sensitivity of the four NDE methods to be directly compared. The results indicated that the acousto-ultrasonic method will not meet requirements for accuracy, inspection time, and reproducibility, while the vibrothermography method will not meet requirements for integration within remanufacturing cells and process flows. The IR methods, thermal wave imaging and flash thermography, both exhibited acceptable accuracy and reproducibility; however, the former method does not meet the requirement for inspection time. Flash IR thermography met all requirements in the initial round of testing, and that result was confirmed in a second round of testing on samples with a larger, more complex geometry. A final recommendation for further development of flash IR thermography was therefore made based upon i) detection performance, and ii) feasibility of deployment into the target application of resurfacing CAT engine blocks and headers.
This repository contains data and processing scripts necessary to train the object detection models utilized in the underwater target detection software demonstration on the RivGen turbine project and to produce performance metrics (precision, recall, mAP50, mAP50-95). - Contents - Data consist of "images" and "labels". Each image has an associated label, both share the same time string in its file name (e.g., 2024_05_25_09_01_57.98.jpg and 2024_05_25_09_01_57.98.txt). Time strings have the format %yyyy_%mm_%dd_%HH_%MM_%SS.%3f. Images and labels were curated from 2021 and 2024 smolt outmigration periods at the project site in Igiugig, AK. Images are monochrome 8-bit images of objects (smolt, debris, and other) passing through the field of view of the deployed cameras during various operational stages of the RivGen turbine. Labels are text files indicating the class and bounding polygon of each object in an image. The provided labels use the "YOLO" label format. - Requirements - Python3.8+ is required to install and run the train and validation script. The README.md provides instruction for installing the requirements from the requirements.py file. - Instructions - The "example_train.py" file ingests the provided data, trains a model, and produces model performance metrics at completion. NOTE: model performance metrics will vary from run to run as a consequence of the random selection of training and validation data.
One of the major scientific goals of the Deep Underground Neutrino Experiment (DUNE) is to detect and measure the neutrino flux originating from galactic core-collapse supernovae. These neutrinos provide an opportunity to study the end of life evolution of massive stars, and reveal information about the structure of core-collapse that is not visible in observations of the electromagnetic spectrum. Because of the rarity of these events, it is crucial that DUNE is able to detect supernova neutrino interactions when they occur. However, this will require sifting through a large quantity of data, motivating the development of a trigger algorithm to identify significant events and discard irrelevant data. Machine learning provides a potential approach to building this trigger. This project generates ADC and ground truth images of simulated neutrino interactions in a LArTPC detector to be used for machine learning, and uses them to train a sparse Convolutional Neural Network (C NN). The performance of this model when applied to the task of pixel classification based on interaction type is examined. This project found that the sparse CNN approach has the potential to have high accuracy in pixel classification, meaning it may be highly relevant to the development of a supernova neutrino trigger for the DUNE far detector.
A method and apparatus for in-situ x-ray scatter imaging of battery electrodes. An apparatus includes an X-ray source, a grid, the grid comprising stainless steel wires with uniform spacing, and a cell, the X-ray source directing a beam of energy through the metal grid and components of the cell, the cell blurring a previously sharp projection of grid wires on an image detector. A method includes providing a Spatial Frequency Heterodyne Imaging system, providing a grid, providing a cell, generating X-rays from the Spatial Frequency Heterodyne Imaging system that pass through components of the cell and the grid, and detecting a scatter image from the X-rays.
PlantCV is an open-source Python project aimed at developing tools to address a range of image-based, plant phenotyping questions. PlantCV has been used for more than 10 years to automate trait collection from image data, and the newest release, PlantCV version 4, continues to lower the barrier to entry for users without substantial coding experience through extensive example use-case tutorials and simplified installation. In addition to usability, we document added functionality since the release of PlantCV v2, including support for more image types such as fluorescence, thermal, and hyperspectral data. Finally, we describe the development of a new subpackage focused on morphological trait measurements like leaf angle, and demonstrate its utility as compared to more manual methods of data collection.
This research program aims to address critical challenges in subsurface exploration and monitoring of anthropogenic CO 2 storage by advancing quantitative dynamic sensing and imaging technologies. Two primary applications—petrophysical assessment of hydrocarbon reservoirs (C1) and CO 2 storage monitoring (C2)—require significant innovation to overcome three barriers (B1-B3). These barriers include the need for enhanced understanding of interactions between physical wave fields (thermoacoustic, electromagnetic, acoustic/seismic, and X-ray) with fluid-filled porous media, the development of multi-sensor data fusion for real-time imaging, and upscaling microscopic quantum effects for macroscopic observations. The project’s objective was to establish a unified 4D coded sensing and imaging approach, integrating EM, AC/S, and TA fields for multi-scale and multi-physics material characterization and subsurface imaging.
During the typical operation of a Concentrating Solar Power (CSP) plant, a large portion of the energy (~45%) can be lost due to various imperfect conditions, such as blocking, shading, mirror soiling, tracking and canting errors, etc. It is necessary to develop efficient and effective field inspection technology to optically survey and characterize a CSP field, which can be used as an input for autonomous control, maintenance scheduling and spot repair whenever necessary to maximize the overall efficiency of the plant. In this project, We aim to apply polarimetric imaging for CSP collector and receiver inspection and develop polarimetric drone cameras (via integrating polarimetric imagers onto drones) for autonomous field inspection in CSP plants.
The project Site Characterization of the Highest-Priority Geologic Formations for CO2 Storage in Wyoming is one of 9 site characterization projects that were implemented as part of ARRA (American Recovery and Reinvestment Act). Data from this project was used to improve resolution of data in NATCARB in the area of study. Data related to this study has already been incorporated in NATCARB Atlas. The Wyoming Carbon Underground Storage Project (WY-CUSP) consisted of CO2 storage site characterization and evaluation, focusing on Wyoming’s most promising CO2 storage reservoirs (the Pennsylvanian Weber/Tensleep Sandstone and Mississippian Madison Limestone) and premier CO2 storage site (Rock Springs Uplift). Results from the WY-CUSP project suggest the two reservoirs could store up to 17,000 million tons of CO2. The WY-CUSP team drilled a stratigraphic test well and acquired a 3-D seismic survey covering 25 square miles of the Rock Springs Uplift site. The team retrieved 916 feet of core from the 12,810-foot-deep well, along with a complete log suite, borehole images, fluid samples, and other data. Project partners (1) provided continuous visual documentation of the core, including grain size, mineralogy, facies distribution, and porosity; (2) performed continuous permeability and velocity scans of selected reservoir intervals; and (3) chemically analyzed the fluid samples. WY-CUSP scientists integrated seismic attributes with observations from log suites, a VSP survey, core, fluid samples, and laboratory analyses, including continuous permeability scans. From these integrations, researchers constructed 3-D spatial distribution volumes of reservoir and seal properties that represent geological heterogeneity at the targeted CO2 storage site. The WY-CUSP team used this data to perform new CO2 plume migration simulations. Baker Hughes, Inc., completed a series of small-scale, in-situ water injectivity measurements. A database was formed when observations, analyses, and experiments from the stratigraphic test well were integrated. Correlation of these data allowed petrophysical parameters to be extrapolated from the test well out into the storage domain (5x5 mile 3-D seismic survey volume). This resulted in an improved, realistic understanding of performance assessments for potential CO2 storage scenarios. The WY-CUSP team worked on (1) improving CO2 storage resource estimates, (2) establishing long-term integrity and permanence of confining layers, (3) designing a profitable strategy for pressure management, and (4) evaluating the utilization of stored CO2 at the Rock Spring Uplift. Finally, Baker Hughes developed a microseismic baseline for the test site using in-bore geophones to complete field operations.
In this article, we describe a modified implementation of Mask Region-based Convolutional Neural Networks (Mask-RCNN) for cosmic ray muon clustering in a liquid argon TPC and applied to MicroBooNE neutrino data. Our implementation of this network, called sMask-RCNN, uses sparse submanifold convolutions to increase processing speed on sparse datasets, and is compared to the original dense version in several metrics. The networks are trained to use wire readout images from the MicroBooNE liquid argon time projection chamber as input and produce individually labeled particle interactions within the image. These outputs are identified as either cosmic ray muon or electron neutrino interactions. We find that sMask-RCNN has an average pixel clustering efficiency of 85.9% compared to the dense network's average pixel clustering efficiency of 89.1%. We demonstrate the ability of sMask-RCNN used in conjunction with MicroBooNE's state-of-the-art Wire-Cell cosmic tagger to veto events containing only cosmic ray muons. The addition of sMask-RCNN to the Wire-Cell cosmic tagger removes 70% of the remaining cosmic ray muon background events at the same electron neutrino event signal efficiency. Furthermore, this event veto can provide 99.7% rejection of cosmic ray-only background events while maintaining an electron neutrino event-level signal efficiency of 80.1%. In addition to cosmic ray muon identification, sMask-RCNN could be used to extract features and identify different particle interaction types in other 3D-tracking detectors.
In this article, we describe a modified implementation of Mask Region-based Convolutional Neural Networks (Mask-RCNN) for cosmic ray muon clustering in a liquid argon TPC and applied to MicroBooNE neutrino data. Our implementation of this network, called sMask-RCNN, uses sparse submanifold convolutions to increase processing speed on sparse datasets, and is compared to the original dense version in several metrics. The networks are trained to use wire readout images from the MicroBooNE liquid argon time projection chamber as input and produce individually labeled particle interactions within the image. These outputs are identified as either cosmic ray muon or electron neutrino interactions. We find that sMask-RCNN has an average pixel clustering efficiency of 85.9% compared to the dense network's average pixel clustering efficiency of 89.1%. We demonstrate the ability of sMask-RCNN used in conjunction with MicroBooNE's state-of-the-art Wire-Cell cosmic tagger to veto events containing only cosmic ray muons. The addition of sMask-RCNN to the Wire-Cell cosmic tagger removes 70% of the remaining cosmic ray muon background events at the same electron neutrino event signal efficiency. This event veto can provide 99.7% rejection of cosmic ray-only background events while maintaining an electron neutrino event-level signal efficiency of 80.1%. In addition to cosmic ray muon identification, sMask-RCNN could be used to extract features and identify different particle interaction types in other 3D-tracking detectors.
This project presents a detailed understanding of the local electronic transport in two different types of conductors. The experimental tools are a combination of local magnetic probes and electronic transport measurements. The first set of samples were superconducting microstructures that exhibited unusual transport behavior. Magnetic imaging revealed the presence of a strain field in the microstructures, which spatially modulated the superconducting properties. This strain field can be controlled through a variety of design parameters opening opportunities for a broad range of quantum materials including topological phases of matter. The second set of samples was fabricated from Cr-doped (Bi,Sb) 2 Te 3 heterostructures that exhibited the quantum anomalous Hall effect (QAHE). Magnetic imaging visualized the current flow in these devices, while simultaneous transport measurements verified that the devices exhibit the QAHE. Surprisingly, the current flowed in the interior of the sample even when exhibiting a high level of quantization.
Clouds play a leading role in the Earth's global energy and solar radiation balance and hydrological cycle. Improving cloud models requires detailed information on the cloud microphysical properties, such as droplet size distribution and number density, liquid water content and cloud composition (droplets, ice particles), which can only be provided by aerial in situ measurements. However, for many atmospheric measurement instruments, the lack of flexibility in selecting the operational mode during operation can lead to uncertainties in sampling and measurement characteristics under continuously varying atmospheric conditions. This SBIR project is developing an advanced, compact optical imaging technology for in situ characterization of cloud hydrometeors. The development involves a deep modification of the existing Mesa Photonics’ Cloud Droplet Measurement System (CDMS) in order to implement real-time automatic adaptive sampling based on the acquired in situ data and environmental parameters. The new system, CDMS-2, implements two measurement modes: side-scatter imaging for smaller hydrometeors and direct bright-field-illumination imaging for larger hydrometeors in a significantly larger sample volume. The system measures the droplet size distribution (DSD) and number density with an added capability of discriminating between liquid water and ice hydrometeors (based on polarization-resolved side-scatter imaging). The instrument will implement automatic switching or alternating between the regular side-scatter imaging mode and sparse/large hydrometeor mode (based on the acquired data). Other adaptive sampling capabilities include variable sample volume and dynamic range (based on the measured DSD). The preferred deployment platforms are uncrewed aircraft systems (UAS) and tethered balloon/kite systems (TBS). The Phase I project achieved (or exceeded) the goals listed in the Work Plan. A CDMS-2 laboratory prototype implementing the polarization-resolved side-scatter imaging mode and direct bright-field-illumination imaging mode was designed and built. Additional capabilities included the variable illumination pulse energy and sample volume. The smallest detectable droplet diameter was improved to 3–4 μm (from the nominal 10 μm value specified for the original CDMS). Discrimination between water droplets and ice particles was experimentally demonstrated. The Phase I prototype was extensively tested and calibrated in the laboratory and also tested in the Pi Cloud Chamber at Michigan Technological University (MTU). The two intensive experimental campaigns at MTU provided unique opportunities of testing the CDMS-2 laboratory prototype under realistic warm and mixed-phase cloud conditions (stable for long periods of time), testing different sampling modes and intercomparing the CDMS-2 prototype to other co-located cloud characterization instruments. The Phase I project successfully demonstrated the feasibility of the proposed technology and identified the engineering challenges of designing a field deployable prototype instrument in Phase II. The Phase I study provides a solid basis for development, characterization and field-testing of the proposed advanced cloud probe with adaptive sampling in Phase II followed by commercialization of the technology in Phase III.
Observation of dynamic processes by transmission electron microscopy (TEM) is an attractive technique to experimentally analyze materials’ nanoscale phenomena and understand the microstructure-properties relationships in nanoscale. Even if spatial and temporal resolutions of real-time TEM increase significantly, it is still difficult to say that the researchers quantitatively evaluate the dynamic behavior of defects. Images in TEM video are a two-dimensional projection of three-dimensional space phenomena, thus missing information must be existed that makes image’s uniquely accurate interpretation challenging. Therefore, even though they are still a clustering high-dimensional data and can be compressed to two-dimensional, conventional statistical methods for analyzing images may not be powerful enough to track nanoscale behavior by removing various artifacts associated with experiment; and automated and unbiased processing tools for such big-data are becoming mission-critical to discover knowledge about unforeseen behavior. We have developed a method to quantitative image analysis framework to resolve these problems, in which machine learning and particle filter estimation are uniquely combined. The quantitative and automated measurement of the dislocation velocity in an Fe-31Mn-3Al-3Si autunitic steel subjected to the tensile deformation was performed to validate the framework, and an intermittent motion of the dislocations was quantitatively analyzed. The framework is successfully classifying, identifying and tracking nanoscale objects; these are not able to be accurately implemented by the conventional mean-path based analysis.
The outskirts of galaxies have been studied from multiple perspectives for the past few decades. However, it is still unknown if all galaxies have clear-cut edges similar to everyday objects. We address this question by developing physically motivated criteria to define the edges of galaxies. Based on the gas density threshold required for star formation, we define the edge of a galaxy as the outermost radial location associated with a significant drop in either past or ongoing in situ star formation. We explore ~1000 low-inclination galaxies with a wide range in morphology (dwarfs to ellipticals) and stellar mass (10 7 M ⊙ < M * < 10 12 M ⊙ ). The location of the edges of these galaxies (R edge ) were visually identified as the outermost cutoff or truncation in their radial profiles using deep multi-band optical imaging from the IAC Stripe82 Legacy Project. We find this characteristic feature at the following mean stellar mass density, which varies with galaxy morphology: 2.9 ± 0.10 M ⊙ pc -2 for ellipticals, 1.1 ± 0.04 M ⊙ pc -2 for spirals, and 0.6 ± 0.03 M ⊙ pc -2 for present-day star-forming dwarfs. Additionally, we find that R edge depends on its age (colour) where bluer galaxies have larger R edge at a fixed stellar mass. The resulting stellar mass–size plane using R edge as a physically motivated galaxy size measure has a very narrow intrinsic scatter (≲0.06 dex). These results highlight the importance of new deep imaging surveys to explore the growth of galaxies and trace the limits of star formation in their outskirts.
Imaging nanoscale radiation-induced defects using the transmission electron microscope (TEM) is a key factor in the successful implementation of materials for nuclear energy structural applications. Analyzing each defect in a TEM micrograph is currently a manual task. To identify the defects in a single image can take anywhere from 15 min to an hour and a project can require the analysis of anywhere from tens to ≥ 100 images. Here, we use artificial intelligence (AI) models to automate this task. For simplification, we evaluated images with only a single type of defect; helium bubbles. Additionally, we performed semantic segmentation of these helium bubble defects in electron microscopy images of irradiated FeCrAl alloys using a deep learning DefectSegNet model. This model, which was previously used to classify crystal defects, is inspired by the classic DenseNet and U-Net image segmentation models. It claims high spatial resolution, but has poor performance at object boundaries. Our paper improves the DefectSegNet model’s application by adding two new features. First, the DefectSegNet model is applied not only to perform calculation pixel-wise but also object (or feature) wise. Because object-wise metrics are directly relevant to our final goal of detecting bubbles, whereas pixel-wise classification is only an intermediate step, it is an important part of our study. Second, a distance map loss (DML) function has been added to increase its performance at object boundaries. It is crucial to accurately represent defects boundaries, especially bubbles, in order to track the bubble-induced swelling caused by irradiation. The boundary-focused DML function is also compared to other loss functions like Cross-entropy, Weighted Binary Cross Entropy (WBCE), Dice and Intersection over Union (IOU). Finally, by incorporating new features, we found a marked improvement on segmentation quality and better shape preservation at the boundaries and areas of the bubbles.
Despite the fact that most microorganisms occupy two- and three-dimensional space in heterogeneous arrangements and in proximity to other microorganisms of different species, much of our current knowledge about their metabolic processes is derived from homogenizing, then analyzing, the chemical content of cultures containing a single microbial species. This project addressed this fundamental dichotomy by developing a set of optical imaging principles combining multiple new technologies. The imaging strategies developed in this project combine metaoptics structures with active in situ nanoscale control of the chemical environment and applied them to a microbial system, Myxococcus xanthus, with particular relevance to the DoE mission. The combination of metaoptical architectures and nanoscale control over the molecular environment enables: (a) precise control over the electromagnetic (EM) field at length scales smaller than the wavelength of light; (b) control of the interaction of the EM field with critical molecular systems in DOE relevant microbes; (c) control over the chemical environment – especially the presence and quantity of reactive oxygen species (ROS) that can affect redox homeostasis; and (d) the ability to ask new kinds of questions not accessible to ‘omics’ approaches or standard methods of biological imaging. These capabilities are applicable to detailed studies of metabolic pathways in microbes and to lignocellulosic biomass deconstruction. To accomplish these objectives, we pursued two over-arching technical goals: (1) the development of new metaoptics-enabled approaches to imaging and spectroscopic characterization; and (2) the development of tools to control the chemical environment of a microbial sample with nanometer-scale precision. Goal 1 was addressed through the design, fabrication, and characterization of new metasurfaces capable of super-resolution imaging through extreme confinement of the optical field. Goal 2 was addressed by controlling the redox potential on the nanoscale in microbial communities and characterizing their effect on intrinsic bacterial fluorophores which act as molecular sentinels and through characterization of soluble factors secreted by Myxococcus xanthus by confocal Raman imaging. The optical imaging/sensing approaches developed here make it possible to use these powerful new imaging and sensing modalities in metabolic studies by making it possible to visualize and track the spatial and temporal expression patterns of natural or engineered pathways in microorganisms.