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.
From quantum chemical and experimental NMR data, a 3D graph neural network, CASCADE, has been developed to predict carbon and proton chemical shifts. Stereoisomers and conformers of organic molecules can be correctly distinguished.
Designing lipid nanoparticle (LNP) delivery systems with specific targeting, potency and minimal side effects is crucial for their clinical use. However, traditional characterization methods, such as dynamic light scattering, cannot accurately quantify physicochemical properties of LNPs and how these are influenced by the lipid composition and mixing method. Here, we structurally characterize polydisperse LNP formulations by applying emerging solution-based biophysical methods that have higher resolution and provide biophysical data beyond size and polydispersity. These techniques include sedimentation velocity analytical ultracentrifugation, field-flow fractionation followed by multiangle light scattering and size-exclusion chromatography in line with synchrotron small-angle X-ray scattering. Here, we show that LNPs have intrinsic polydispersity in size, RNA loading and shape, which depend on both the formulation technique and the lipid composition. Lastly, we predict LNP transfection in vitro and in vivo by examining the relationship between mRNA translation and physicochemical characteristics. Solution-based biophysical methods will be essential for determining LNP structure–function relationships, facilitating the creation of new design rules for LNPs.
The proposed Direct Feed High Level Waste (DFHLW) approach for processing high-level tank waste at Hanford is intended to reduce processing time by bypassing the Pretreatment Facility and transferring waste directly from the tank farm to the WTP HLW vitrification facility. This processing strategy could reduce or eliminate the washing and leaching steps that would have occurred in the Pretreatment facility. Operation of the vitrification facility is subject to chemical and radiological limits protecting safety (e.g. Waste Acceptance Criteria, or WACs) and process quality (e.g. Process Control Limits, or PCLs). Without washing and leaching, there is a greater risk of exceeding the WACs and PCLs. Hanford process engineers have devised blending strategies based on known chemical and radiological composition, volumes, and solids loadings of individual layers within each waste tank. These blending campaigns succeed in predicting a processing strategy that does not exceed the WACs and PCLs. However, the calculations do not ascribe uncertainties to the tank analysis data, quantities of material taken from the tanks to make the blend, or potential for mixing of layers within tanks. In order to confirm that a process strategy is working, it would be advantageous to have inline or at-line analytical instrumentation installed in the processing facilities that deliver measurement results in real time.
The Geothermal Operational Optimization with Machine Learning (GOOML) project has developed a generic and extensible component-based system modeling framework to study complex geothermal fields using a data-driven approach. Through building a digital twin of a geothermal steam field with the GOOML modeling framework, operators can analyze historical and forecasted power production, explore possible steam field configurations, and optimize real world operations, all in a cost-effective digital environment. The GOOML modeling software is based on a historical data-assimilation framework that uses first-principal thermodynamics to model steam field components using historical data, and a forecast framework that uses machine-learning-driven models of steam field components to predict future operations. This modeling framework creates countless new opportunities for digital exploration of steam field design and operations. To date, digital twins have been developed for several steam fields in New Zealand and the United States. These digital twins have been validated by comparing hindcast predictions against historical production data. Field design and operations have been explored using genetic optimization and reinforcement learning. Initial results show compelling and often surprising opportunities for improved design and operation of fields with 2 to 5 percent improvements in annual energy production. GOOML is driving a step-change in geothermal operations by applying state-of-the-art machine learning algorithms, comprehensive data analytics, and a first-of-its-kind intelligent geothermal systems model.
Yttrium hydride is an optimal choice of moderator material for thermal neutron spectrum reactors requiring small core volumes. The Transformation Challenge Reactor (TCR) program at Oak Ridge National Laboratory (ORNL) aims to develop an additively manufactured microreactor core by leveraging recent advances in materials, manufacturing, data analytics, and high-fidelity modeling and simulation. Yttrium hydride was selected as the moderator material for TCR due to its high moderating power, which enables the spatially efficient reactor design and thermal stability that is superior when compared with zirconium hydride. Hydrogen desorption from yttrium hydride is expected at elevated temperatures, posing safety and performance concerns. As a lifetime component in advanced reactors, appropriate strategies to mitigate hydrogen release from yttrium hydride over long-term reactor operation are required. Understanding hydrogen mobility in yttrium hydride provides the scientific basis for developing a hydrogen barrier for hydride moderator and predicting the hydrogen redistribution within the material during various operational conditions. This study investigated the hydrogen motion in YH 1.87 as a function of temperature ranging from 200 to 1,173 K using incoherent quasielastic neutron scattering (IQNS) at the Spallation Neutron Source (SNS). The results are presented and discussed in terms of hydrogen self-diffusion coefficients, activation energy for translational hydrogen motion, and hydrogen jump distances. Based on the IQNS data, YH 1.87 is found to be stable at high temperatures with no translational H motion below 1,023 K within the IQNS detection limits. In the temperature range of 1,073–1,173 K, hydrogen diffusivity in YH 1.87 is $D = 4.57 X 10^{⁻3} [m^2/s] exp (⁻\frac{1.73[eV]}{RT}$). The hydrogen concentration’s dependence on hydrogen diffusivity is also discussed in this report.
Potential performance gains from optimal (non-causal) impedance-matching control of wave energy devices in irregular ocean waves are dependent on deterministic wave elevation prediction techniques that work well in practical applications. Although a number of devices are designed for operation in intermediate water depths, little work has been reported on deterministic wave prediction in such depths. Here, we investigated a deterministic wave-prediction technique based on an approximate propagation model that leads to an analytical formulation, which may be convenient to implement in practice. To improve accuracy, an approach to combine predictions based on multiple up-wave measurement points is evaluated. The overall method is tested using experimental time-series measurements recorded in the U.S. Navy MASK basin in Carderock, MD, USA. For comparison, an alternative prediction approach based on Fourier coefficients is also tested with the same data. Comparison of prediction approaches with direct measurements suggest room for improvement. Possible sources of error including tank reflections are estimated, and potential mitigation approaches are discussed.
Grain boundaries are critically important to the material performance of fusion reactor materials such as vanadium, particularly mechanical properties and irradiation resistance. A key challenge to the design and control of grain boundaries in vanadium alloys is the lack of quantitative data on grain boundary segregation. In this study, we combine computational and experimental methods to address this gap. Furthermore, using a machine learning-accelerated quantum mechanics/molecular mechanics approach, we calculated the segregation spectra for 28 transition metal elements in polycrystalline vanadium, and validated these predictions experimentally for a subset of solutes that sample a range of segregation behavior, specifically zirconium, titanium, and tungsten, using analytical transmission electron microscopy. Furthermore, the agreement between experiment and theory highlights the predictive capability of our approach. Critically, this work provides a comprehensive database of quantum-accurate solute segregation enthalpies in vanadium, enabling the development of advanced alloys for fusion reactors applications.
We propose a general machine learning-based framework for building an accurate and widely applicable energy functional within the framework of generalized Kohn–Sham density functional theory. To this end, we develop a way of training self-consistent models that are capable of taking large datasets from different systems and different kinds of labels. Here, we demonstrate that the functional that results from this training procedure gives chemically accurate predictions on energy, force, dipole, and electron density for a large class of molecules. It can be continuously improved when more and more data are available.
We developed models to predict the photoactivity of organic molecules for photocatalytic hydrogen evolution by integrating experiment, computation, and machine learning. This marks a step toward the data-driven discovery of molecular photocatalysts.
This report documents the experimental conditions and results for the testing of the Mirion Technologies CAAS-3S Criticality Accident Alarm System (CAAS) for the Y-12 National Security Complex (Y-12) and Uranium Processing Facility (UPF). The testing was conducted at the Godiva IV Burst Reactor at the National Criticality Experiments Research Center (NCERC) at the Nevada National Security Site (NNSS) during the week of January 11, 2021. The testing was conducted in accordance with 25774-QL-POA-JR00-00002-VDE-27.0, CAAS-3S Radiation Testing of Cabinet Components Test Plan. The tests subjected the CAAS-3S system to intense and short duration mixed neutron and gamma radiation fields and to high integrated radiation doses. The purpose of the test was to environmentally qualify various CAAS-3S components in a radiation environment. Successful completion of the tests was predicated on the collaborative efforts of several organizations. Sandia National Laboratory (SNL) and Lawrence Livermore National Laboratory (LLNL) provided dosimetry to estimate photon and neutron dose. Los Alamos National Laboratory (LANL) manages the NCERC facility, operated the Godiva IV reactor, and provided field support for CAAS-3S placement and communications interface for data acquisition systems supplied by Mirion Technologies. Throughout the test campaign, Y-12 and UPF personnel provided technical support in determining changes to the test schedule, as needed, interpretation of the results, and comparison between the measured dosimetry data and predicted results. Throughout the test campaign, Mirion personnel provided technical support in the operation of the CAAS-3S system itself.
Drilling in the offshore environment involves a complex network of infrastructure including pipelines, platforms, rigs, subsea installations, ports, and terminals. Government and industry partners have developed this network over many decades and it remains a critical part of the United States (U.S.) energy portfolio. Many of the major components of this system have been designed with a 20- to 30-year lifespan, yet consistent and growing energy demands support the need to extend the design life of existing infrastructure or repurpose it for secondary needs (i.e. enhanced oil recovery, carbon storage, and new wells). As a result, a growing portion of the offshore infrastructure in the U.S. is approaching or has exceeded its original design life. A critical step in ensuring the continued safe and effective operation of offshore infrastructure is developing a comprehensive understanding of the state of offshore infrastructure and the factors that effect it. The purpose of this project is to assess the current state of existing infrastructure and identify the factors involved in infrastructure degradation through the development and application of big data analytics, machine learning, and advanced spatio-temporal analysis. The project leverages existing data at NETL and combines it with new information on offshore oil and gas structures and the ambient offshore environment in an effort to identify patterns associated with infrastructure integrity. Building on the identified trends and patterns, this project incorporates exploratory analytics and spatial analysis tools in conjunction with machine learning and statistical models to characterize the condition of existing platforms in the offshore environment and predict their risk of failure.
Arsenene, a less-explored two-dimensional material, holds the potential for applications in wearable electronics, memory devices, and quantum systems. This study introduces a bond-order potential model with Tersoff formalism, the ML-Tersoff, which leverages multireward hierarchical reinforcement learning (RL), trained on an ab initio data set. This data set covers a spectrum of properties for arsenene polymorphs, enhancing our understanding of its mechanical and thermal behaviors without the complexities of traditional models requiring multiple parameter sets. Our RL strategy utilizes decision trees coupled with a hierarchical reward strategy to accelerate convergence in high-dimensional continuous search spaces. Unlike the Stillinger-Weber approach, which demands separate formalisms for buckled and puckered forms, the ML-Tersoff model concurrently captures multiple properties of the two polymorphs by effectively representing the local environment, thereby avoiding the need for different atomic types. Here, we apply the ML model to understand the mechanical and thermal properties of the arsenene polymorphs and nanostructures. We observe an inverse relationship between the critical strain and temperature in arsenene. Thermal conductivity calculations in nanosheets show good agreement with ab initio data, reflecting a decrease in thermal conductivity attributable to increased anharmonic effects at higher temperatures. We also apply the model to predict the thermal behavior of arsenene nanotubes.
Site specific microstructure control is a critical research area within the field of additive manufacturing due to its potential to revolutionize part performance. One way to achieve site specific microstructure control is through control of the solidification conditions via the construction of intricate scan paths; however, the search space for such a problem is large. Previous attempts only considered the solidification conditions at the top surface while also requiring either lots of manual-fine tuning or large amounts of computational resources. This paper introduces a general method for scan path optimization which considers the solidification conditions in the bulk of the material without an increase in computational expense. This method consists of three core components:1. A heat transfer model for simulating the temperature field at a given time.2. A surrogate model which takes scan pattern information and temperature data and predicts the solidification conditions of the bulk as well as the meltpool depths for a spot melt.3. A decision algorithm to decide which spot melt should be printed next based on the outputs of the surrogate model.Each of these components can be changed without changing the overall method. Within this work, this method is applied in the creation of an algorithm containing a semi-analytic heat transfer model to simulate the temperature field, a fully convolutional neural network (FCNN) as the surrogate model, and a greedy decision algorithm. The resulting algorithm produced complex scan patterns which gave strong results for simulated microstructure control.
In this article, we develop a generalized and scalable statistical model for accurate prediction of fouling resistance using commonly measured parameters of industrial heat exchangers. This prediction model is based on deep learning where a scalable algorithmic architecture learns non-linear functional relationships between a set of target and predictor variables from large number of training samples. Here, the efficacy of this modeling approach is demonstrated for predicting fouling in an analytically modeled cross-flow heat exchanger, designed for waste heat recovery from flue-gas using room temperature water. The performance results of the trained models demonstrate that the mean absolute prediction errors are under 10 –4 KW –1 for flue-gas side, water side and overall fouling resistances. The coefficients of determination (R 2 ), which characterize the goodness of fit between the predictions and observed data, are over 99%. Even under varying levels of measurement noise in the inputs, we demonstrate that predictions over an ensemble of multiple neural networks achieves better accuracy and robustness to noise. We find that the proposed deep-learning fouling prediction framework learns to follow heat exchanger flow and heat transfer physics, which we confirm using locally interpretable model agnostic explanations around randomly selected operating points. Overall, we provide a robust algorithmic framework for fouling prediction that can be generalized and scaled to various types of industrial heat exchangers.
Abstract We present a data-driven approach to characterizing nonidentifiability of a model’s parameters and illustrate it through dynamic as well as steady kinetic models. By employing Diffusion Maps and their extensions, we discover the minimal combinations of parameters required to characterize the output behavior of a chemical system: a set of effective parameters for the model. Furthermore, we introduce and use a Conformal Autoencoder Neural Network technique, as well as a kernel-based Jointly Smooth Function technique, to disentangle the redundant parameter combinations that do not affect the output behavior from the ones that do. We discuss the interpretability of our data-driven effective parameters, and demonstrate the utility of the approach both for behavior prediction and parameter estimation. In the latter task, it becomes important to describe level sets in parameter space that are consistent with a particular output behavior. We validate our approach on a model of multisite phosphorylation, where a reduced set of effective parameters (nonlinear combinations of the physical ones) has previously been established analytically.
We present angular diameter measurements obtained by measuring the position of baryon acoustic oscillations (BAO) in an optimized sample of galaxies from the first three years of Dark Energy Survey data (DES Y3). The sample consists of 7 million galaxies distributed over a footprint of 4100 deg 2 with 0.6 < z photo < 1.1 and a typical redshift uncertainty of 0.03 ( 1 + z ) . The sample selection is the same as in the BAO measurement with the first year of DES data, but the analysis presented here uses three times the area, extends to higher redshift, and makes a number of improvements, including a fully analytical BAO template, the use of covariances from both theory and simulations, and an extensive preunblinding protocol. We used two different statistics; angular correlation function and power spectrum, and validate our pipeline with an ensemble of over 1500 realistic simulations. Both statistics yield compatible results. We combine the likelihoods derived from angular correlations and spherical harmonics to constrain the ratio of comoving angular diameter distance D M at the effective redshift of our sample to the sound horizon scale at the drag epoch. We obtain D M ( z eff = 0.835 ) / r d = 18.92 ± 0.51 , which is consistent with, but smaller than, the Planck prediction assuming flat Λ CDM , at the level of 2.3 σ . The analysis was performed blind and is robust to changes in a number of analysis choices. It represents the most precise BAO distance measurement from imaging data to date, and is competitive with the latest transverse ones from spectroscopic samples at z > 0.75 . When combined with DES 3 x 2 pt + SNIa , they lead to improvements in H 0 and Ω m constraints by ∼ 20 % .
The Transformational Challenge Reactor (TCR) program is leveraging additive manufacturing (AM) technologies to fabricate nuclear components which will be assembled into a fully functional microreactor core. Compared with traditional manufacturing technologies, AM technologies allow (1) real-time observation of the manufacturing process at a much higher resolution using in-situ monitoring technologies to capture the sensor signatures that scientifically describe each event occurring over time and space and (2) validation of the manufacturing process quality using domain-informed data analytics techniques as a potential qualification and certification methodology for the final component. This report provides an update on the program work on in-situ and ex-situ data correlation and associated data analytics results. Examples are provided to illustrate progress with respect to laser powder bed fusion (L-PBF), binder jetting, computed tomography (CT) reconstruction, and mechanical testing. Elements of the Digital Thread and data management infrastructure are discussed in the main document, and an extensive supplemental appendix is provided detailing the Digital Platform, as well as its implementation and subcomponents. In conclusion, the path forward for the next fiscal year is also discussed.
Herein, data-driven approaches are currently renovating the field of heterogenous catalysis and open a door to advance catalyst design. Their success depends heavily on the synergy among machine learning (ML), experimental data, and quantum mechanical (QM) calculations. In this brief survey of recent progress, we examine catalysis informatics in the context of (1) ML-aided catalyst characterizations, (2) knowledge extractions from experimental data, (3) predictions of catalytic properties and constructions of reaction networks, and (4) ML-enabled large-scale QM simulations. An outlook on the current challenges of this rapidly evolving field is also provided.