SEARCH · Engineering Papers
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.
A neural network for segmenting and processing Lorentz TEM images of skyrmion lattices. This software based on U-Net architecture is capable of identifying skyrmions in out-of-focus Lorentz TEM images based on training data. This isspecifically useful for analyzing large area data from in-situ experiments.
The project developed quantitative convergent-beam electron diffraction (QCBED) methods by energy-filtered transmission electron microscopy (EFTEM) and used them in combination with density functional theory (DFT) calculations to study the electron density distribution in metallic and intermetallic phases with different cubic and non-cubic crystal structures that comprise elements with d-electron shells. The experimental methods developed here focus on the bonding charge distribution as one of the quantum mechanical characteristics central for understanding of intrinsic properties and validation of DFT calculations. Multiple structure and temperature factors have been measured simultaneously from nano-scale volumes of high-quality crystal with sufficient accuracy and precision for comparison with electron density distribution calculations by DFT. The often anisotropic temperature factors for the different atoms and atom sites in chemically ordered phases can differ significantly from those known for relevant pure element crystals due to bonding effects. Thus they have been measured from the same crystal volumes from which the structure factors have been determined. The ferromagnetic ordered intermetallic phases FePd and FePt are selected as model systems for 3d-4d and 3d-5d electron interactions, while the intermetallic phases NiAl and TiAl are used to probe 3d-3p electron interactions. Additionally, pure transition metal elements with d-electrons have been studied. FCC metals exhibit well defined delocalized bonding charge in tetrahedral sites, while less directional, more distributed bonding charge attains in BCC metals. Agreement between DFT calculated and QCBED results degrades as d-electron levels fill in the elements, and for intermetallics as d-d interactions become prominent over p-d interactions. Utilizing the LDA+U approach enabled inclusion of onsite Coulomb-repulsion effects in DFT calculations, which can afford improved agreements with QCBED results. This implies that onsite Coulomb repulsion effects become non-negligible as the d-orbitals fill. The use of now easily measured low-order structure factors as an additional experimental metric in validation of DFT calculation of electronic structures of crystals offers potential to capture better both total energy related properties and details of the interatomic bonding in system with d-electron orbital contributions. This effort advanced the state of the art in quantitative TEM experimentation, provides original experimental data uniquely suited for new validation approaches of DFT calculations of d-electron affected transition metals and intermetallics.
Three-dimensional finite strain and magnetic susceptibility anisotropy have been determined for 15 ordinary chondrites. The axes of strain and magnetic ellipsoids roughly correlate in both magnitude and orientation. The shapes of these ellipsoids are generally oblate spheroids that define a dominant foliation and a weak lineation. These characteristics suggest deformation involving uniaxial compaction. The degree of uniaxial deformation correlates with intensity of shock, as indicated by optical, TEM and chemical criteria. These data, plus the lack of a relationship between foliation and metamorphic history, indicate that dynamic processes, i.e., impacts, produced planar deformation fabrics in chondrites.
Standardisation of data collection and analysis is essential to enable commercialisation of 2D materials in a wide range of technologies. Selected area electron diffraction (SAED) in the transmission electron microscope (TEM) is one of the key methods for distinguishing monolayer from bilayer and few-layer graphene by comparing the 1st and 2nd order diffraction spot intensities. Yet there are many factors that can affect the reliability of data collection and interpretation, causing the measurement of monolayer samples to deviate from the literature boundary condition of $I_{\{\bar{2}110\}}$$/$$I_{\{1\bar{1}00\}}$ < 1 for monolayer graphene (1LG). Here we present the results of a large interlaboratory SAED comparison study, where 15 international laboratories measured and analysed nominally identical samples of chemical vapour deposited graphene. Large variations were observed in the measured ratios of diffraction spot intensities, with the largest variance associated with poor quality SAED data resulting from inadequate specimen handling and storage. To inform the reliable determination of monolayer thickness from SAED patterns we provide a description of best practice for specimen handling, TEM operation, data collection and analysis. This work was undertaken within VAMAS Technical Working Area 41: Graphene and related 2D materials—Project 9, the results of which have been directly incorporated into ISO/TS 21356–2 for the characterisation of graphene sheets. We find that when this methodology is followed, 1LG can be distinguished from bilayer or thicker material with high confidence where analysis of a single SAED pattern gives $I_{\{\bar{2}110\}}$$/$$I_{\{1\bar{1}00\}}$ < 1.2, even in the absence of precise specimen tilting.
In computed TEM tomography, image segmentation represents one of the most basic tasks with implications not only for 3D volume visualization, but more importantly for quantitative 3D analysis. In case of large and complex 3D data sets, segmentation can be an extremely difficult and laborious task, and thus has been one of the biggest hurdles for comprehensive 3D analysis. Heterogeneous catalysts have complex surface and bulk structures, and often sparse distribution of catalytic particles with relatively poor intrinsic contrast, which possess a unique challenge for image segmentation, including the current state-of-the-art deep learning methods. To tackle this problem, we apply a deep learning-based approach for the multi-class semantic segmentation of a γ-Alumina/Pt catalytic material in a class imbalance situation. Specifically, we used the weighted focal loss as a loss function and attached it to the U-Net’s fully convolutional network architecture. We assessed the accuracy of our results using Dice similarity coefficient (DSC), recall, precision, and Hausdorff distance (HD) metrics on the overlap between the ground-truth and predicted segmentations. Our adopted U-Net model with the weighted focal loss function achieved an average DSC score of 0.96 ± 0.003 in the γ-Alumina support material and 0.84 ± 0.03 in the Pt NPs segmentation tasks. We report an average boundary-overlap error of less than 2 nm at the 90th percentile of HD for γ-Alumina and Pt NPs segmentations. The complex surface morphology of γ-Alumina and its relation to the Pt NPs were visualized in 3D by the deep learning-assisted automatic segmentation of a large data set of high-angle annular dark-field (HAADF) scanning transmission electron microscopy (STEM) tomography reconstructions.
Here, compressive sensing (CS) is a fast-emerging approach for improving data throughput by taking advantage of the inherent compressibility of real world data—the same kind of compressibility that allows millions of users to stream full-screen video without breaking the internet. Transmission electron microscope (TEM) users regularly collect multi-GB data sets and then use aggressive data reduction strategies to draw out the tiny residual of Shannon entropy that answers the question they wanted to ask. Much of the raw-byte-count information in an acquisition is random, redundant, or irrelevant, yet its collection still represents substantial cost in terms of instrumentation bandwidth and sample damage.
This project addresses the needs of the U.S. Department of Energy (DOE) to develop advanced monitoring technologies and protocols to track the fate of subsurface carbon dioxide (CO2) plumes for carbon storage. Specifically, the project seeks to develop and test a unique and novel system of technologies consisting of electromagnetic data acquisition, coupled multiphysics imaging, and reservoir model enhancement to understand the migration and long-term distribution of CO2 in the subsurface. The overarching objective is to develop an integrated approach for long term monitoring of carbon storage. The two main components of the project include the methodology development and the test of the method at a field site. The methodology component consists of 1) developing the field procedure and protocol for collecting time-lapse controlled-source electromagnetic (CSEM) data with source electric current injected into the subsurface through wellbore casings; 2) building of background 3D electrical conductivity utilizing multiple sources of data such as supplemental surface transient EM (TEM) surveys, well-logs, and seismic structural information, for enhancing CSEM signal from reservoir depths; and 3) coupled multiphysics simulations and inversion of CSEM data constrained by production data and by structural information from seismic imaging of the reservoir and overlying formations. The testing component used the field site of Bell Creek Oil Field, which served both as a field laboratory for the method development as well as a test site to evaluate the CSEM signal strengths and the methodology developed in this research project. We have accomplished all the proposed tasks and developed the methodology as planned. These include the procedure for time-lapse CSEM data acquisition, data processing techniques, integration with 3D conductivity model building, fast reservoir simulation for history matching using machine learning, and interpreting CSEM data with coupling to the reservoir modeling. Collectively, the outcome of these tasks form a coherent workflow that can be applied to monitor dedicated carbon storage in saline reservoirs. The testing component evaluated the applicability and limitations of the method, and concluded that the method would be ideal for monitoring dedicated carbon storage sites utilizing saline reservoirs.
Acquiring atomic-resolution images of beam sensitive materials or under dynamic conditions (e.g. operando STEM) is very difficult and requires precise dose fractioning to minimize the dose delivered to the sample and acquire the data more rapidly than dynamics present. Typical methods of low-dose STEM imaging include adjusting the beam current, dwell-time, and probe Scanning transmission electron microscopy (STEM) is used routinely to obtain atomic-resolution images of a small selection of radiation tolerant materials under stable conditions.size. All three of these determine the dose over a spatial area in STEM. Recently, compressive sensing [1] has been applied to electron microscopy. The computational imaging method of inpainting [2] has been proposed as a method to reduce dose and increase acquisition speed for STEM. Inpainting is a technique to recover missing pixels in an image. By skipping pixels (randomly) during acquisition, a fourth dose-control mechanism is realized [3]. A few implementations of randomly-subsampled STEM exist (e.g., [4]). To make the concepts of compressive sensing accessible to the microscopy community we begin with an example using atomic resolution images of a GaAs crystal (see Figure 1). Because the crystal is periodic it has a sparse representation in reciprocal space. In fact, a sparse representation is a key component of compressive sensing. If an image is sub-sampled randomly we find that the Fourier transform remains sparse, but becomes increasing noisy as less and less data is acquired. Yet, by finding the peaks in the reciprocal image and removing the noise it is possible to recover atomic resolution images with only 1% of the pixels. Figure 1 shows the reconstruction process for GaAs sub-sampled at 50%, 10%, 5%, and 1%. Periodic images form a special class for which the optimal sampling approach (according to CS theory) is a random selection of pixels. For non-periodic images, such as those arising in in-situ STEM, methods such as dictionary learning are necessary [2,5]. The data can be compressed in time for TEM [5] or pixels can be skipped randomly for STEM. There are several issues that arise in developing a random scan for STEM [4] and a coded aperture for TEM. The remainder of the talk will address these issues. References: [1] E Candes and M Wakin. Signal Processing Magazine, IEEE. 2(25) (2008), pp. 21. [2] A Stevens, Pu Y, Sun Y, et al. Artificial Intelligence & Statistics (AISTATS), 2017. [3] A Stevens, H Yang, L Carin et al. Microscopy 63(1), (2014), pp. 41. [4] L Kovarik, A Stevens, A Liyu et al. Appl. Phys. Lett. 109, 164102 (2016). [5] A Stevens, L Kovarik, P Abellan et al. Adv. Structural and Chemical Imaging 1(10), (2015), pp. 1. [6] Supported by the Chemical Imaging, Signature Discovery, and Analytics in Motion Initiatives at PNNL. PNNL is operated by Battelle Memorial Inst. for the US DOE; contract DE-AC05-76RL01830.
This article is a Commentary on McMurtrie & Näsholm et al., 218: 119–130. Roots are represented in Terrestrial Ecosystem Models (TEMs) in much less detail than their equivalent above-ground resource acquisition organs – leaves. Often roots in TEMs are simply resource sinks, and below-ground resource acquisition is commonly simulated without any relationship to root dynamics at all, though there are exceptions (e.g. Zaehle & Friend, 2010). The representation of roots as carbon (C) and nitrogen (N) sinks without complementary source functions can lead to strange sensitivities in a model. For example, reducing root lifespans in the Community Land Model (version 4.5) increases plant production as N cycles more rapidly through the ecosystem without loss of plant function (D. M. Ricciuto, unpublished). The primary reasons for the poorer representation of roots compared with leaves in TEMs are three-fold: (1) data are much harder won, especially in the field; (2) no simple mechanistic models of root function are available; and (3) scaling root function from an individual root to a root system lags behind methods of scaling leaf function to a canopy. Here in this issue of New Phytologist, McMurtrie & Näsholm (pp. 119–130) develop a relatively simple model for root N uptake that mechanistically accounts for processes of N supply (mineralization and transport by diffusion and mass flow) and N demand (root uptake and microbial immobilization).
Big Data has already proven itself as a valuable tool that lets geographers and urban researchers utilize large data re- sources to generate new insights. However, wider adoption of Big Data techniques in these areas is impeded by a number of difficulties in both knowledge discovery and data and sci- ence production. Typically users face such problems as dis- parate and scattered data, data management, spatial search- ing, insufficient computational capacity for data-driven anal- ysis and modelling, and the lack of tools to quickly visual- ize and summarize large data and analysis results. Here we propose an architecture for an Urban Information Sys- tem (UrbIS) that mitigates these problems by utilizing the Big Data as a Service (BDaaS) concept. With technolog- ical roots in High-performance Computing (HPC), BDaaS is based on the idea of outsourcing computations to dif- ferent computing paradigms, scalable to super-computers. UrbIS aims to incorporate federated metadata search, in- tegrated modeling and analysis, and geovisualization into a single seamless workflow. The system is under active devel- opment and is built around various emerging technologies that include hybrid and NoSQL databases, massively par- allel systems, GPGPU computing, and WebGL-based geo- graphic visualization. UrbIS is designed to facilitate using Big Data across multiple cities to better understand how ur- ban areas impact the environment and how climate change and other environmental change impact urban areas.
Neural networks are promising tools for high-throughput and accurate transmission electron microscopy (TEM) analysis of nanomaterials, but are known to generalize poorly on data that is “out-of-distribution” from their training data. Given the limited set of image features typically seen in high-resolution TEM imaging, it is unclear which images are considered out-of-distribution from others. Here, we investigate how the choice of metadata features in the training dataset influences neural network performance, focusing on the example task of nanoparticle segmentation. We train and validate neural networks across curated, experimentally collected high-resolution TEM image datasets of nanoparticles under various imaging and material parameters, including magnification, dosage, nanoparticle diameter, and nanoparticle material. Overall, we find that our neural networks are not robust across microscope parameters, but do generalize across certain sample parameters. Additionally, data preprocessing can have unintended consequences on neural network generalization. Our results highlight the need to understand how dataset features affect deployment of data-driven algorithms.
The atomic structure of Ni-Pd nanoparticles has been studied using atomic pair distribution function (PDF) analysis of X-ray total scattering data and with transmission electron microscopy (TEM). Larger nanoparticles have PDFs corresponding to the bulk face-centered cubic packing. However, the smallest nanoparticles have PDFs that strongly resemble those obtained from bulk metallic glasses (BMGs). In fact, by simply scaling the distance axis by the mean metallic radius, the curves may be collapsed onto each other and onto the PDF from a metallic glass sample. In common with a wide range of BMG materials, the intermediate range order may be fit with a damped single-frequency sine wave. When viewed in high-resolution TEM, these nanoparticles exhibit atomic fringes typical of those seen in small metallic clusters with icosahedral or decahedral order. These two seemingly contradictory results are reconciled by calculating the PDFs of models of icosahedra that would be consistent with the fringes seen in TEM. These model PDFs resemble the measured ones when significant atom-position disorder is introduced, drawing together the two diverse fields of metallic nanoparticles and BMGs and supporting the view that BMGs may contain significant icosahedral or decahedral order.
Monodisperse carbon nanodots and nanodiamond particles were studied by AFM method. Particle size distributions were calculated using AFM and DLS data. The results obtained correspond to TEM measurements. It was demonstrated that AFM allows detecting nanoparticles with sizes less than 2 nm. The average size of the particles for both samples was found to be ~4 nm. It was found that PSD for CND and DND was symmetric and asymmetric, correspondingly. Such difference can be understood in terms of different crystal structure perfection and the specific aspects of nanosized carbon preparation technology.
Within the 200 East Area of the Central Plateau and southeastward toward the Columbia River, a high-hydraulic conductivity zone (HCZ) has been interpreted to extend through the 200-PO-1 and 200-BP-5 operable units on the Hanford Site. The HCZ is a controlling hydraulic feature that impacts groundwater flow out of the 200 East Area and the fate of eastwardly migrating plumes from the 200 West Area. The lateral extent of the HCZ is highly uncertain, and despite strong evidence for the existence of the HCZ based on water-level data and contaminant plume tracking, there is still a limited understanding of how to define its boundaries. To provide additional information on the nature and extent of the HCZ, three surface geophysical methods – electrical resistivity tomography (ERT), time-domain electromagnetics (TEM), and seismic methods – were used to collect data south of 200 East. In addition, existing data from 200 East, consisting of surface seismic data, a borehole check shot survey in 699-37-47A, and borehole stratigraphic interpretations, were used to aid interpretations of newly collected seismic data south of 200 East. This work presumed that the contrast in subsurface geophysical properties would be a first-order aid identifying a transmissive zone(s) within the HCZ analysis area by imaging contrasts and/or anomalies in geophysical properties. While seismic, ERT, and TEM methods have sensitivity to overlapping physical properties (porosity, moisture content, lithology), the resolution and physics used to acquire each of these datasets are different, and therefore the information can also be different. Figure S.1 shows the locations of the geophysical data considered in this report.
Neuromorphic computing stands as a promising frontier for advancing AI algorithms and applications like ChatGBT, offering significant energy efficiency gains. This paper delves into the hardware design intricacies of memristive thin films and their elementary switching mechanisms, including anion migration, electron migration, and phase transitions. Through comprehensive analysis of electron energy loss spectroscopy (EELS) data via in-situ transmission electron microscopy (TEM), we will deduce the primary memristive switching mechanisms vital for optimizing thin film fabrication parameters and achieving desired film thickness, conductivity, and memory retention. A single crystal ptype Si substrate was used with TiN as the bottom metal electrode, TiO x as the insulating dielectric layer, and Pt as the top metal electrode. In-situ TEM was able to tell us the thin film didn’t behave like a filamentary or phase transition material. EELS data deduced that electron trapping/detrapping was one of the primary switching mechanisms. By shedding light on these elementary mechanisms, our study aims to catalyze the development of more 2 efficient and effective neuromorphic computing systems to be deployed into mainstream technologies.
Time-of-flight neutron diffraction measurements were conducted at ambient conditions to study microstructures of the δ-phase 239 Pu-2at% Ga alloy. Based on the line profile analysis of diffraction data we derived dislocation densities using a correction routine to account for the anisotropic strain broadening. Our results show here that the average dislocation density in the as-received sample, previously treated at cryogenic temperatures, is five times higher than the dislocation density measured after annealing the sample, indicating that the neutron instrumentation used in this work has sufficient resolution to detect the dislocation density changes in different sample processing. These results are also in reasonably good agreement with the dislocation densities previously reported based on TEM observations and x-ray diffraction data, suggesting that the simple correction routines applied in this work have a fairly good fidelity for deriving important microstructural information such as dislocation density for elastically highly anisotropic δ-phase 239 Pu-Ga alloys.
Highlights: • C. pneumoniae infection promotes monocyte TEM through increasing VEC permeability. • C. pneumoniae infection stimulates VE-cadherin tyrosine phosphorylation by Src. • VE-cadherin phosphorylation is involved in VE-cadherin internalization. • VE-cadherin phosphorylation is involved in EC hyperpermeability caused by infection. • VE-cadherin phosphorylation is associated with the infection-induced monocyte TEM. Migration of monocytes into the subendothelial layer of the intima is one of the critical events in early atherosclerosis. Chlamydia pneumoniae (C. pneumoniae) infection has been shown to promote monocyte transendothelial migration (TEM). However, the exact mechanisms have not yet been fully clarified. In this study, we tested the hypothesis that C. pneumoniae infection increases vascular endothelial cell (VEC) permeability and subsequent monocyte TEM through stimulating the tyrosine phosphorylation of vascular endothelial-cadherin (VE-cadherin). Here, we demonstrated that C. pneumoniae infection promoted monocyte TEM in a TEM assay possibly by increasing the permeability of a VEC line EA.hy926 cell as assessed by measuring the passage of FITC-BSA across a VEC monolayer. Subsequently, Western blot analysis showed that C. pneumoniae infection induced VE-cadherin internalization. Our further data revealed that Src-mediated VE-cadherin phosphorylation at Tyr658 was involved in C. pneumoniae infection-induced internalization of VE-cadherin, VEC hyperpermeability and monocyte TEM. Taken together, our data indicate that C. pneumoniae infection promotes monocyte TEM by increasing VEC permeability via the tyrosine phosphorylation and internalization of VE-cadherin in VECs.
In-core nuclear materials including fuel pins and cladding materials fail due to issues including corrosion, mechanical wear, and pellet cladding interaction. In most such scenario microstructure dependent and corrosioninduced chemistry dependent property changes significantly affect performance of cladding, pellet, and housing. Emphasis of this work was on replace conventional pellet-cladding material models with a new straingradient viscoplasticity model that is informed by transmission electron microscopy (TEM) based measurements and by nanomechanical Raman spectroscopy (NMRS) based measurements. The TEM measurements are quantitative in nature and therefore reveal stress-strain relations with simultaneous insights into mechanisms of deformation at nanoscale. The NMRS measurements reveal the similar information at mesoscale along with additional information on relating local microstructural stresses with applied stresses. The resulting information is used to fit constants in the strain gradient viscoplasticity model as well as to validate one. During TEM measurements, a micro-electro-mechanical system based setup was developed with mechanical actuation, sensing, heating, and electrical loading. Contrary to post-mortem analysis or qualitative visualization, this setup combines direct visualization of the mechanisms behind deformation with measurement of stress, strain, thermal and electrical properties. The unique research philosophy of visualizing the microstructure at high resolution while measuring the properties led to fundamental understanding in grain size and temperature effects on measured mechanical properties such as fracture toughness. A key contribution is the role of mechanical loading boundary conditions to deconvolute the insitu TEM based nanoscale and NMRS based mesoscale data to bulk behavior. First the literature based pellet cladding mechanical interaction model based on the work of Retel’s and Williamson’s in literature work to predict tempurature and stress distribution in cladding and pellet at normal operating condition was analyzed. Later the data was fitted to find constants for a viscoplastic strain gradient model. The developed model still needs to be refined and calibrated using various experimental results. That remains the focus of future work. Overall, a major thrust of the work was therefore on active control of the microstructure (grain size, defect density and types) exploiting the multi-physics coupling in materials. In particular, using experiments the synergy of current density, mechanical stress and temperature were studied to annihilate defects and recrystallize grains. The developed model is being examined for implementation in BISON. Multiple invited talks, international journal publications, and conference publications were performed by students supported on this work. Another output is support multiple PhD and masters thesis students who will be an important asset for future basic nuclear research. Future Work Recommendations: A nuclear reactor operates under significant variations of thermal loads due to energy cycling and mechanical loads due to constraint effects. Significant thermal and chemical diffusion takes place at the pallet-cladding level. While the proposed work established new experimental approach and new dataset for Zircaloy-4, the irradiation level was in the range of 1-2 dpa. Samples with higher dpa need to be examined. Therefore, a continual of support of the performed work is essential. Currently, these are the only experiments that can measure the produced data. The work also needs to be extended to different fuel types and cladding types such as SiC and FeCrAl based claddings. A combination of datasets for these materials can then be used to analyze accurately predict behavior of critical pellet cladding systems in accident scenario with high heat flux and high thermal loads. This is a BIG unknown as if now.