Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “tomography problems”

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 145 records · Page 8
All sourcesNASA NTRSDOE OSTI
DOE OSTI2022

Data-centric machine learning in quantum information science

Abstract We propose a series of data-centric heuristics for improving the performance of machine learning systems when applied to problems in quantum information science. In particular, we consider how systematic engineering of training sets can significantly enhance the accuracy of pre-trained neural networks used for quantum state reconstruction without altering the underlying architecture. We find that it is not always optimal to engineer training sets to exactly match the expected distribution of a target scenario, and instead, performance can be further improved by biasing the training set to be slightly more mixed than the target. This is due to the heterogeneity in the number of free variables required to describe states of different purity, and as a result, overall accuracy of the network improves when training sets of a fixed size focus on states with the least constrained free variables. For further clarity, we also include a ‘toy model’ demonstration of how spurious correlations can inadvertently enter synthetic data sets used for training, how the performance of systems trained with these correlations can degrade dramatically, and how the inclusion of even relatively few counterexamples can effectively remedy such problems.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2024

Four-dimensional phase space tomography from one-dimensional measurements of a hadron beam

In this paper, we use one-dimensional measurements to infer the four-dimensional phase space density of an accumulated proton beam in the Spallation Neutron Source (SNS) accelerator. The reconstruction was performed by maximizing the distribution’s entropy subject to the measurement constraints and thus represents the most conservative inference from the data. The reconstructed distribution reproduces the measured profiles down to the noise level, and simulations indicate that the problem is reasonably well constrained. Similar measurements could serve as benchmarks for beam dynamics simulations in the SNS or hadron accelerators.

43 PARTICLE ACCELERATORS↗
DOE OSTI2019

Microstructure and Properties of A356 Alloy Wheels Fabricated by Low-Pressure Die Casting with Local Squeeze

The low-pressure die-casting process (LPDC) is widely used in fabricating aluminum alloy automotive wheels. With good castability and mechanical properties, A356 alloys are important and common materials for LPDC wheels. However, in practical casting process, defects often appear, especially porosities in the wheel thick-walled parts (i.e., spoke) due to inadequate interdendritic feeding, which greatly decreases the mechanical properties. In this investigation, to deal with this problem, local squeeze (LS) was added in conventional LPDC to enhance the interdendritic feeding. The practical A356 alloy wheel castings were fabricated with different LS velocities and times. The quantitative characteristics of α-Al cells, Si particles, and porosities with and without the LS analyzed by optical microscope, scanning electron microscope, and 3D x-ray computed tomography were compared. Their effects on the tensile and impact properties were discussed. The results have shown that the LS can increase the cooling rate and feed shrinkage of spoke parts, hence resulting in finer grains and denser microstructure. In result, the properties of spoke parts were improved with the LS.

36 MATERIALS SCIENCE↗
DOE OSTI2023

Resolution-enhanced X-ray fluorescence microscopy via deep residual networks

Abstract Multimodal hard X-ray scanning probe microscopy has been extensively used to study functional materials providing multiple contrast mechanisms. For instance, combining ptychography with X-ray fluorescence (XRF) microscopy reveals structural and chemical properties simultaneously. While ptychography can achieve diffraction-limited spatial resolution, the resolution of XRF is limited by the X-ray probe size. Here, we develop a machine learning (ML) model to overcome this problem by decoupling the impact of the X-ray probe from the XRF signal. The enhanced spatial resolution was observed for both simulated and experimental XRF data, showing superior performance over the state-of-the-art scanning XRF method with different nano-sized X-ray probes. Enhanced spatial resolutions were also observed for the accompanying XRF tomography reconstructions. Using this probe profile deconvolution with the proposed ML solution to enhance the spatial resolution of XRF microscopy will be broadly applicable across both functional materials and biological imaging with XRF and other related application areas.

36 MATERIALS SCIENCE↗
NASA NTRS1990

Using data with models - Ill-posed problems

An attempt is made to show that the combination of data with models leads to a sequence of conventionally ill-posed problems which can be treated in a systematic and practical way by control-theory techniques. The illustrative approach is to exploit the very great mathematical simplifications which result when the methods of finite dimensional vector spaces are used, thus reducing a lot of difficult mathematics to classical least squares. The immediate motivation for this study comes from ocean acoustic tomography.

Wunsch, C.↗
DOE OSTI2019

Accelerating full waveform inversion via source stacking and cross-correlations

SUMMARY Accurate synthetic seismic wavefields can now be computed in 3-D earth models using the spectral element method (SEM), which helps improve resolution in full waveform global tomography. However, computational costs are still a challenge. These costs can be reduced by implementing a source stacking method, in which multiple earthquake sources are simultaneously triggered in only one teleseismic SEM simulation. One drawback of this approach is the perceived loss of resolution at depth, in particular because high-amplitude fundamental mode surface waves dominate the summed waveforms, without the possibility of windowing and weighting as in conventional waveform tomography. This can be addressed by redefining the cost-function and computing the cross-correlation wavefield between pairs of stations before each inversion iteration. While the Green’s function between the two stations is not reconstructed as well as in the case of ambient noise tomography, where sources are distributed more uniformly around the globe, this is not a drawback, since the same processing is applied to the 3-D synthetics and to the data, and the source parameters are known to a good approximation. By doing so, we can separate time windows with large energy arrivals corresponding to fundamental mode surface waves. This opens the possibility of designing a weighting scheme to bring out the contribution of overtones and body waves. It also makes it possible to balance the contributions of frequently sampled paths versus rarely sampled ones, as in more conventional tomography. Here we present the results of proof of concept testing of such an approach for a synthetic 3-component long period waveform data set (periods longer than 60 s), computed for 273 globally distributed events in a simple toy 3-D radially anisotropic upper mantle model which contains shear wave anomalies at different scales. We compare the results of inversion of 10 000 s long stacked time-series, starting from a 1-D model, using source stacked waveforms and station-pair cross-correlations of these stacked waveforms in the definition of the cost function. We compute the gradient and the Hessian using normal mode perturbation theory, which avoids the problem of cross-talk encountered when forming the gradient using an adjoint approach. We perform inversions with and without realistic noise added and show that the model can be recovered equally well using one or the other cost function. The proposed approach is computationally very efficient. While application to more realistic synthetic data sets is beyond the scope of this paper, as well as to real data, since that requires additional steps to account for such issues as missing data, we illustrate how this methodology can help inform first order questions such as model resolution in the presence of noise, and trade-offs between different physical parameters (anisotropy, attenuation, crustal structure, etc.) that would be computationally very costly to address adequately, when using conventional full waveform tomography based on single-event wavefield computations.

Romanowicz, Barbara↗
DOE OSTI2018

Roll Compaction of Powdered Radioactive Wastes - 18251

Most of low- and intermediate-level radioactive wastes must be solidified by a proper binding matrix and disposed at the centralized disposal site. So it is desirable to achieve a high loading of radwastes in a drum to reduce the disposal fee while meeting both the national regulations and the waste acceptance criteria of the disposal site. In general, the radioactive wastes (liquid, dried powder and sludge) were solidified by mixing with a proper binding matrix (cement, polymer, asphalt, etc.), but their solidified volume inevitably increases due to the limitation of incorporating of radwaste into a binding matrix. To solve this problem, some Korean NPPs operate both a CTS (Concentrate Treatment System, evaporation and granulation) and a polymer solidification system. The size distribution of granulated wastes produced by CTS is so broad (diameter; several μm ∼ several cm) that a polymer cannot impregnate the void space between particles due to their small size. However, if the powdered radwaste can be compacted into high density pellets, and the pellets can be loaded in a drum with maximum waste loading, then the binding matrix can fill the void between pellets, it might be more desirable to keep the size distribution of pellets constant to reduce the solidified waste volume. This compaction methodology can greatly reduce the disposal fee due to the reduction of the solidified waste volume, and can facilitate polymer solidification due to the constant size of the radwaste pellets. This methodology is also applicable to solidify powdered wastes such a depleted uranium used as a catalyst, ash and sludge. Loading radwaste compacted with bentonite pellets versus particles into a container serves, as an engineered or natural barrier in a deep geologic disposal system. Roll compaction is a typical agglomeration process for dry powders and depends on the device parameters and the material properties. This study, firstly, was focused on the design of roll compactor (briquetting machine), the shape and size of pellet, and secondly the loading method of pellets into a drum, their solidification in polymer, and evaluating the overall volume reduction. We used bentonite (particle size; 85 ∼ 100 μm) as a material to be compacted instead of radwaste powders (concentrates and depleted uranium wastes) without the addition of agglomeration agents. In the design of roll, we considered the diameter (200 mm) and width (41 mm) of roll, two types of pocket shapes (< 9 mm, 0.275 ml), and the surface of the pocket. Fig. 2 shows two pellet shapes, tetragonal and circular. The pressure of roll compactor, one of the main process variables, is > 320 kg f/cm{sup 2}, and the roll gap is 0.5 mm. The powders are compacted in a compaction zone between two counter-rotating rolls according to the complicated mechanisms. As a result of the state and the apparent density of the pellets obtained by changing the pressure from 0 to 286 kg f/cm{sup 2}, the pellets were unstable at pressure=0, but above 148 kg f/cm{sup 2} the pellets were more stable, and the apparent density of the pellets was 2.63 g/cm{sup 3} at 148 kg f/cm{sup 3}, and 2.68 g/cm{sup 3} at 286 kg f/cm{sup 2}. The roller speed (rpm) can determine the capacity (kg/h) of a compactor, and depends on both the compaction pressure and the feeding rate. By increasing the roll speed from 2 to 4 rpm, the apparent density of pellets was decreased gradually from 2.74 g/cm{sup 3} to 1.45 g/cm{sup 3}. We designed the special feeding system in order to feed powders uniformly to the roll gripping zone and to supply powders into the narrowest part of the roll gap. At a constant roll speed (3 rpm) and a constant compaction pressure (286 kg f/cm{sup 2}), the apparent density of the pellets was linearly proportional to the feeding rate from 2.18 to 2.74 g/cm{sup 3} by changing the feeding rate of powders from 8 to 12 rpm. In the polymer solidification of the obtained pellets, we used epoxy resin as a solidification agent, and showed their solidified waste forms in Fig. 5. The incorporation ratio (wt.%) of powders/polymer was about 71.7/28.3 in tetragonal pellets, 71.9/28.1 in circular pellets. We calculated the volume reduction in an aspect of powders itself and the solidified waste form. In powder case, through the comparison of the apparent density of the obtained pellets (2.74 g/cm{sup 3}) and raw powders (0.93 g/cm{sup 3}), the volume reduction was calculated as about 2.74/0.93 = 2.95 times. This means that powders can be compacted to 1/2.95 times. And in waste form case, the weight ratio of powder/polymer in a mixing method (workable) and a compaction method is about 0.42, 2.56, individually. Therefore the volume reduction was calculated as about 2.56/0.42 = 6.01 times. We found a compaction methodology for the volume reduction of powdered radioactive wastes, and their polymer solidification will be more effective to reduce the volume of radwastes to be disposed. After a characterization of the solidified waste form incorporated pellets (including sintered pellets), this methodology will be applied to treat depleted uranium wastes used as catalyst, and to solve the problem of NPPs' polymer solidification system, as mentioned above. (authors)

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗
DOE OSTI2025

Tomographic Sparse View Selection Using the View Covariance Loss

Standard computed tomography (CT) reconstruction algorithms such as filtered back projection (FBP) and Feldkamp-Davis-Kress (FDK) require many views for producing high-quality reconstructions, which can slow image acquisition and increase cost in non-destructive evaluation (NDE) applications. Over the past 20 years, a variety of methods have been developed for computing high-quality CT reconstructions from sparse views. However, the problem of how to select the best views for CT reconstruction remains open. In this paper, we present a novel view covariance loss (VCL) function that measures the joint information of a set of views by approximating the normalized mean squared error (NMSE) of the reconstruction. We present fast algorithms for computing the VCL along with an algorithm for selecting a subset of views that approximately minimizes its value. Our experiments on simulated and measured data indicate that for a fixed number of views our proposed view covariance loss selection (VCLS) algorithm results in reconstructions with lower NRMSE, fewer artifacts, and greater accuracy than current alternative approaches.

Lin, Jingsong [Purdue University]↗
DOE OSTI2022

Information content of and the ability to reconstruct dichroic X-ray tomography and laminography

Dichroic tomography is a 3D imaging technique in which the polarization of the incident beam is used to induce contrast due to the magnetization or orientation of a sample. The aim is to reconstruct not only the optical density but the dichroism of the sample. The theory of dichroic tomographic and laminographic imaging in the parallel-beam case is discussed as well as the problem of reconstruction of the sample’s optical properties. The set of projections resulting from a single tomographic/laminographic measurement is not sufficient to reconstruct the magnetic moment for magnetic circular dichroism unless additional constraints are applied or data are taken at two or more tilt angles. For linear dichroism, three polarizations at a common tilt angle are insufficient for unconstrained reconstruction. However, if one of the measurements is done at a different tilt angle than the other, or the measurements are done at a common polarization but at three distinct tilt angles, then there is enough information to reconstruct without constraints. Possible means of applying constraints are discussed. Furthermore, it is shown that for linear dichroism, the basic assumption that the absorption through a ray path is the integral of the absorption coefficient, defined on the volume of the sample, along the ray path, is not correct when dichroism or birefringence is strong. This assumption is fundamental to tomographic methods. An iterative algorithm for reconstruction of linear dichroism is demonstrated on simulated data.

Marcus, Matthew A. (ORCID:0000000325277586)↗
DOE OSTI2022

A deep learning approach for semantic segmentation of unbalanced data in electron tomography of catalytic materials

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.

36 MATERIALS SCIENCE↗
NASA NTRS2010

Electrical Capacitance Volume Tomography with High-Contrast Dielectrics

The Electrical Capacitance Volume Tomography (ECVT) system has been designed to complement the tools created to sense the presence of water in nonconductive spacecraft materials, by helping to not only find the approximate location of moisture but also its quantity and depth. The ECVT system has been created for use with a new image reconstruction algorithm capable of imaging high-contrast dielectric distributions. Rather than relying solely on mutual capacitance readings as is done in traditional electrical capacitance tomography applications, this method reconstructs high-resolution images using only the self-capacitance measurements. The image reconstruction method assumes that the material under inspection consists of a binary dielectric distribution, with either a high relative dielectric value representing the water or a low dielectric value for the background material. By constraining the unknown dielectric material to one of two values, the inverse math problem that must be solved to generate the image is no longer ill-determined. The image resolution becomes limited only by the accuracy and resolution of the measurement circuitry. Images were reconstructed using this method with both synthetic and real data acquired using an aluminum structure inserted at different positions within the sensing region. The cuboid geometry of the system has two parallel planes of 16 conductors arranged in a 4 4 pattern. The electrode geometry consists of parallel planes of copper conductors, connected through custom-built switch electronics, to a commercially available capacitance to digital converter. The figure shows two 4 4 arrays of electrodes milled from square sections of copper-clad circuit-board material and mounted on two pieces of glass-filled plastic backing, which were cut to approximately square shapes, 10 cm on a side. Each electrode is placed on 2.0-cm centers. The parallel arrays were mounted with the electrode arrays approximately 3 cm apart. The open ends were surrounded by a metal guard to reduce the sensitivity of the electrodes to outside interference and to help maintain the spacing between the arrays. Other uses for this innovation potentially include quantifying the amount of commodity remaining in the fuel and oxidizer tanks while on-orbit without having to fire spacecraft engines. Another orbit application is moisture sensing in plant-growth experiments because microgravity causes moisture in soil to distribute itself in unusual ways. At the moment, the hardware and image reconstruction technique may only be of interest to people involved in nondestructive evaluation. The reconstructed image takes almost a full week to reproduce with existing computer power. However, because computer power and speeds follows Moore s Law, execution times are likely to become acceptable within the next five to eight years. The code was written in Mathematica for dedicated use with the ECVT system. In its present form, it is not suitable to be used directly as a consumer product. However, the code could be likely improved by rewriting it in a compiled language such as C or Fortran.

Nurge, Mark↗
NASA NTRS2004

Passive Acoustic Tomography Tested for Measuring Gas Temperatures

The requirements of higher performance, better fuel economy, and lower emissions place an increasing premium on knowing the internal operating parameters of jet engines. One of the most important is the gas temperature in the post combustor section of the engine. Typically the gas temperature is measured with a thermocouple probe or by some optical technique such as Rayleigh scattering. Probes, while providing valuable information, have several limitations. The probe signal must be corrected for radiation and conduction losses, probes provide only a point measurement, and probes must be constructed of materials whose melting points are lower than the temperature of the environment into which they are inserted. Some of the disadvantages of probes are overcome by various optical techniques. Nothing needs to be inserted into the flow, and the temperature can be directly related to the signal by known physical laws. However, optical techniques require optical access (i.e., a window) and a light source (such as a laser), and they are very sensitive to the presence of particles in the flow. To overcome these problems, researchers from the NASA Glenn Research Center and The University of Nevada are developing a technique that uses sound instead of light to measure gas temperature. Like optical techniques, it is nonintrusive--no probe need be exposed to the combustion environment--and the temperature is directly related to a measured quantity--the speed of sound, which is proportional to the square root of the absolute temperature. The temperature profile inside the engine is constructed from the differences in arrival time between correlated signals from an array of microphones placed around the circumference of the engine. In much the same way as a complete picture of the inside of your body can be constructed from an array of x-ray photographs taken at different angles, the temperature profile in the engine is constructed from the angular array of microphones. It is tomography by sound waves. Active acoustic tomography, in which a sound pulse is injected into the flow and the time delays between members of an array of microphones are used to construct the temperature field has been used successfully in the stacks of power plants. However, the flow field inside a jet engine is much too noisy for it to be possible to detect an externally injected sound pulse. Instead we are developing passive acoustic tomography, which uses the sound already present in the flow.

Fralick, Gustave C.↗
DOE OSTI2023

Machine learning for detection of 3D features using sparse x-ray tomographic reconstruction

In many inertial confinement fusion (ICF) experiments, the neutron yield and other parameters cannot be completely accounted for with one and two dimensional models. This discrepancy suggests that there are three dimensional effects that may be significant. Sources of these effects include defects in the shells and defects in shell interfaces, the fill tube of the capsule, and the joint feature in double shell targets. Due to their ability to penetrate materials, x rays are used to capture the internal structure of objects. Methods such as computational tomography use x-ray radiographs from hundreds of projections, in order to reconstruct a three dimensional model of the object. In experimental environments, such as the National Ignition Facility and Omega-60, the availability of these views is scarce, and in many cases only consists of a single line of sight. Mathematical reconstruction of a 3D object from sparse views is an ill-posed inverse problem. These types of problems are typically solved by utilizing prior information. Neural networks have been used for the task of 3D reconstruction as they are capable of encoding and leveraging this prior information. We utilize half a dozen, different convolutional neural networks to produce different 3D representations of ICF implosions from the experimental data. Deep supervision is utilized to train a neural network to produce high-resolution reconstructions. These representations are used to track 3D features of the capsules, such as the ablator, inner shell, and the joint between shell hemispheres. Machine learning, supplemented by different priors, is a promising method for 3D reconstructions in ICF and x-ray radiography, in general.

Wolfe, Bradley T. (ORCID:0000000268301614)↗
DOE OSTI2024

Enhancing Biopreparedness through a Model System to Understand the Molecular Mechanisms that Lead to Pathogenesis and Disease Transmission: NW-BRaVE

The science of biopreparedness to counter biological threats hinges on understanding the fundamental principles and molecular mechanisms that lead to pathogenesis and disease transmission. Our vision to address this challenge is to create a powerful and user-friendly platform to elucidate the fundamental principles of how molecular interactions drive pathogen-host relationships and host shifts. We will enable groundbreaking discoveries by integrating a wide range of structural, genomics, proteomics, and other advanced omics measurements, along with evolutionary and artificial intelligence predictions. To make sure the system is applicable to real-world problems, we will develop it in the context of a tractable model system, the small, abundant, and accessible photosynthetic cyanobacteria and their constantly co-adapting viral pathogens, cyanophages. This model will maintain the system’s applicability to real-world problems and techniques, but the overall focus will be on elucidating general principles of detecting, assessing, and surveilling molecular interaction, adaptation, and coevolution that are system agnostic and therefore extensible to other viral-host interactions. Our overall objectives are to (1) identify the molecular complexes that comprise the cyanobacteria redox macromolecular subsystem and how they dynamically change with bacteriophage infection in situ, using cryo-electron tomography; (2) profile regulatory changes during infection using proteomics, multiomics, and experimental validation, and integrate the data with in situ structures; (3) use genomics and metagenomics to determine environmental and population factors across time scales that impact the interactions between marine cyanobacteria and their cyanophage parasites, predicting the evolutionary origins of in situ structural and functional interactions, convergence and coevolution; and (4) develop a data integration and transformation platform that facilitates the integration of in situ, proteomic, and evolutionary measurements of molecular interactions to surveil diverse hosts and parasites in various environmental contexts. These objectives address Focus Area 2 Reveal Molecular Interactions Across Biological Scales for Design of Targeted Interventions. Our powerful and user-friendly platform will enhance connections between the often-siloed fields of structure, molecular phenotype, and evolutionary genomics that are key to biopreparedness, but in need of integration (Figure 1). We will build an integrated navigation tool to facilitate the effective use of globally distributed experimental data for integrated analysis and predictive modeling. The project will develop, implement, and test a platform to assess host-pathogen molecular interactions, adaptation to hosts and host shifts, and coevolution between hosts and pathogens, successfully impacting the research community by revolutionizing abilities to study any host-pathogen interaction, encourage diverse community contributions, and gain fundamental insights into how proteins adapt to new contexts. This ability will be critical for designing early interventions to address future threats. We will build surveillance training capability, aiming for a fair and equitable response to future pandemics and biothreats.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2019

Imaging in the pre-operative staging of ovarian cancer

The main prognostic factor in ovarian cancer is the stage of disease at diagnosis. The staging system in use (FIGO classification, updated in 2014) is based on the surgical-pathological findings. Although surgical staging is the gold standard in ovarian cancer, the initial patient management depends on the imaging-based pre-surgical staging assessment, in order to identify unresectable or difficult to resect disease. Radiologists need to be aware of the strengths of the available imaging modalities, as well as the imaging pitfalls. Clear understanding of pattern of disease spread and review areas are critical for accurate staging and treatment planning. The current standard of care for pre-surgical staging is CT of the thorax, abdomen, and pelvis. This allows a rapid evaluation of disease extent and is fairly accurate in identifying bulky disease but has definite limitations in assessing the extent of small volume disease and in the confirmation of certain sites of disease beyond the abdomen. Functional MRI has been reported to be superior in detecting small peritoneal deposits. PET/CT may be used as a problem-solving tool in some patients where determination remains unclear, particularly in confirmation of advanced stage beyond the abdomen.

62 RADIOLOGY AND NUCLEAR MEDICINE↗
DOE OSTI2022

Artificial Intelligence for Accelerating Nuclear Applications, Science, and Technology

Artificial intelligence (AI) and machine learning (ML) methods have had significant impacts in science and technology in recent years. These methods for generating models from datasets or logic-based algorithms that emulate aspects of human performance can similarly accelerate the fields of nuclear applications, science, and technology toward the IAEA goals of contributing to peace, health, and prosperity. In order to accomplish advances with AI in general and ML in particular across these fields, IAEA can play a significant role by establishing, hosting and curating centralised resources, including databases, adhering to FAIR (findable, accessible, interoperable and reusable) principles and Open Science best practices, providing stewardship of data sharing, supporting training efforts and development of relevant workforces, as well as enabling connections among the scientific, technology, mathematics, AI and ethics communities. Many areas can benefit from the use of AI in the realm of nuclear applications. In human health, these areas include clinical research, epidemiology, nutrition, medical imaging, radiotherapy and education of health professionals. AI-based tools are also being used to facilitate different clinical tasks in imaging, computer-assisted diagnosis in mammography and lung cancer screening programmes, and dose prediction in nuclear medicine procedures. ML methods in particular may also increase the efficiency and accuracy of the analysis of computerised tomography and dual-energy absorptiometry scans for body composition and bone analysis. The application of AI methods to nuclear and related technologies in food and agriculture can lead to significant advances and improved efficiency in the optimisation of agricultural production, food product development, management of supply chains, food safety and food authenticity control. In the water and environmental sector, AI can help inform policies to mitigate the world’s water problems. The application of AI techniques to hydrology and environmental sciences is expected to improve patterns identification and enable model predictions under a changing climate.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
NASA NTRS1994

Assessment of impact damage of composite rocket motor cases

This contract reviewed the available literature on mechanisms of low velocity impact damage in filament wound rocket motor cases, MDE methods to quantify damage, critical coupon level test methods, manufacturing and material process variables and empirical and analytical modeling off impact damage. The critical design properties for rocket motor cases are biaxial hoop and axial tensile strength. Low velocity impact damage is insidious because it can create serious nonvisible damage at very low impact velocities. In thick rocket motor cases the prevalent low velocity impact damage is fiber fracture and matrix cracking adjacent to the front face. In contrast, low velocity loading of thin wall cylinders induces flexure, depending on span length and the flexure induces delamination and tensile cracking on the back face wall opposed to impact occurs due to flexural stresses imposed by impact loading. Important NDE methods for rocket motor cases are non-contacting methods that allow inspection from one side. Among these are vibrothermography, and pulse-echo methods based on acoustic-ultrasonic methods. High resolution techniques such as x-ray computed tomography appear to have merit for accurate geometrical characterization of local damage to support development of analytical models of micromechanics. The challenge of coupon level testing is to reproduce the biaxial stress state that the full scale article experiences, and to determine how to scale the composite structure to model full sized behavior. Biaxial tensile testing has been performed by uniaxially tensile loading internally pressurized cylinders. This is experimentally difficult due to gripping problems and pressure containment. Much prior work focused on uniaxial tensile testing of model filament wound cylinders. Interpretation of the results of some studies is complicated by the fact that the fabrication process did not duplicate full scale manufacturing. It is difficult to scale results from testing subscale cylinders since there are significant differences in out time of the resins relative to full scale cylinder fabrication, differences in hoop fiber tensioning and unsatisfactory coupon configurations. It appears that development of a new test method for subscale cylinders is merited. Damage tolerance may be improved by material optimization that uses fiber treatments and matrix modifications to control the fiber matrix interface bonding. It is difficult to develop process optimization in subscale cylinders without also modeling the longer out times resins experience in full scale testing. A major breakthrough in characterizing the effect of impact damage on residual strength, and understanding how to scale results of subscale evaluations, will be a sound micromechanical model that described progressive failure of the composite. Such models will utilize a three dimensional stress analysis due to the complex nature of low velocity impact stresses in thick composites. When these models are coupled with non-contact NDE methods that geometrically characterize the damage and acoustic methods that characterize the effective local elastic properties, accurate assessment of residual strength from impact damage may be possible. Directions for further development are suggested.

Paris, Henry G.↗
DOE OSTI2018

Simulation tools for atom probe tomography: A path for diagnosis and treatment of image degradation

The ideal picture of a near-perfect 3D microscope often presented regarding Atom Probe Tomography faces several issues. These issues degrade the metrological performance of the instrument and find their roots in the phenomena acting at the atomic to the mesoscopic level in the vicinity of the surface of a field emitter. From the field evaporation process at the atomic scale, to the macroscopic scale of the instrument, the path to model the imaging process and to develop more accurate and reliable reconstruction algorithms is not a single lane road. This paper focused on the numerical methods used to understand, treat, and potentially heal imaging issues commonly affecting the data in atom probe experiments. A lot of room for improvement exists in solving accuracy problems observed in complex materials by means of purely electrostatic models describing the image formation in a classical approach. Looking at the sample at the atomic scale, the phenomena perturbing the imaging process are subtle. An examination of atomic scale modifications of the sample surface in the presence of a high surface electric field is therefore mandatory. Atomic scale molecular dynamic models integrating the influence of the high surface electric are being developed with this aim. It is also demonstrated that the complex behavior of atoms and molecules in high fields, and consequences on collected data, can be understood through the use of accurate ab-initio models modified to include the impact of the extreme surface electric field.

77 NANOSCIENCE AND NANOTECHNOLOGY↗