Search NASA⌕ Search

SEARCH · Search NASA

Results for “Task Force”

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 487 records · Page 27
All sourcesNASA NTRSDOE OSTI
DOE OSTI2014

TU-F-18A-09: CT Number Stability Across Patient Sizes Using Virtual-Monoenergetic Dual-Energy CT

Purpose: Virtual-monoenergetic imaging uses dual-energy CT data to synthesize images corresponding to a single photon energy, thereby reducing beam-hardening artifacts. This work evaluated the ability of a commercial virtual-monoenergetic algorithm to achieve stable CT numbers across patient sizes. Methods: Test objects containing a range of iodine and calcium hydroxyapatite concentrations were placed inside 8 torso-shaped water phantoms, ranging in lateral width from 15 to 50 cm, and scanned on a dual-source CT system (Siemens Somatom Force). Single-energy scans were acquired from 70-150 kV in 10 kV increments; dual-energy scans were acquired using 4 energy pairs (low energy: 70, 80, 90, and 100 kV; high energy: 150 kV + 0.6 mm Sn). CTDIvol was matched for all single- and dual-energy scans for a given phantom size. All scans used 128×0.6 mm collimation and were reconstructed with 1-mm thickness at 0.8-mm increment and a medium smooth body kernel. Monoenergetic images were generated using commercial software (syngo Via Dual Energy, VA30). Iodine contrast was calculated as the difference in mean iodine and water CT numbers from respective regions-of-interest in 10 consecutive images. Results: CT numbers remained stable as phantom width varied from 15 to 50 cm for all dual-energy data sets (except for at 50 cm using 70/150Sn due to photon starvation effects). Relative to the 15 cm phantom, iodine contrast was within 5.2% of the 70 keV value for phantom sizes up to 45 cm. At 90/150Sn, photon starvation did not occur at 50 cm, and iodine contrast in the 50-cm phantom was within 1.4% of the 15-cm phantom. Conclusion: Monoenergetic imaging, as implemented in the evaluated commercial system, eliminated the variation in CT numbers due to patient size, and may provide more accurate data for quantitative tasks, including radiation therapy treatment planning. Siemens Healthcare.

07 ISOTOPE AND RADIATION SOURCES↗
DOE OSTI2015

Wigner phase space distribution via classical adiabatic switching

Evaluation of the Wigner phase space density for systems of many degrees of freedom presents an extremely demanding task because of the oscillatory nature of the Fourier-type integral. We propose a simple and efficient, approximate procedure for generating the Wigner distribution that avoids the computational difficulties associated with the Wigner transform. Starting from a suitable zeroth-order Hamiltonian, for which the Wigner density is available (either analytically or numerically), the phase space distribution is propagated in time via classical trajectories, while the perturbation is gradually switched on. According to the classical adiabatic theorem, each trajectory maintains a constant action if the perturbation is switched on infinitely slowly. We show that the adiabatic switching procedure produces the exact Wigner density for harmonic oscillator eigenstates and also for eigenstates of anharmonic Hamiltonians within the Wentzel-Kramers-Brillouin (WKB) approximation. We generalize the approach to finite temperature by introducing a density rescaling factor that depends on the energy of each trajectory. Time-dependent properties are obtained simply by continuing the integration of each trajectory under the full target Hamiltonian. Further, by construction, the generated approximate Wigner distribution is invariant under classical propagation, and thus, thermodynamic properties are strictly preserved. Numerical tests on one-dimensional and dissipative systems indicate that the method produces results in very good agreement with those obtained by full quantum mechanical methods over a wide temperature range. The method is simple and efficient, as it requires no input besides the force fields required for classical trajectory integration, and is ideal for use in quasiclassical trajectory calculations.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2023

A Multi-Objective Bayesian Optimized Human Assessed Multi-Target Generated Spectral Recommender System for Rapid Pareto Discoveries of Material Properties

Optimization for different tasks like material characterization, synthesis, and functional properties for desired applications over multi-dimensional control parameter and function spaces need a rapid strategic search through active learning. However, in all cases prior to optimization, the target material properties are assumed known and fixed, which mostly deviates from real-world scenarios in material synthesis. This can be critical for running expensive experiments on new materials, when the experimental results are fuzzy for any scientific outcomes due to improper target setting, ultimately wasting time and cost. The failure rate and cost are even higher over exploring on multi-target space, where we want to learn the pareto among multiple properties, to jointly optimize during material synthesis for desired applications. To address the challenge, here we introduce the human-operator attempt flexibility in the active learning based automated experiment framework, with generating multiple human assessed targets through a voting-based recommender system during real-time microscope measurements over the large material image space, sequentially learn/update multiple desired targets through a weighting system, and adaptively search in multiple material properties functional space for non-dominated pareto discoveries to maximize the custom structural similarity based acquisition function. We term this a multi-objective Bayesian optimized human assessed multi-target generated spectral recommender systems (MOBO-HAM-SRS). The approach has been demonstrated to peizoresponse force spectroscopy of a ferroelectric thin film, exploring with different kernels and acquisition functions. This work shows an advancement towards human-AI collaborated automated experiments, steering optimization trajectories through human overpowering AI at the early stage when uncertainty is high and AI overpowering human at the later stage with rapid exploration towards optimal goal, following human-assessed multiple targets properties.

Biswas, Arpan↗
DOE OSTI2013

Multipole approximation of special nonlinear magnets for the IOTA ring

All existing focusing optics for accelerators are built to be "linear", i.e. the transverse focusing force is proportional to the particle displacement from the closed orbit. Such approach is used in order to make trajectories stable in the entire area of phase space, because the potential describing linear focusing is integrable and does not produce chaotic motion. Technically, the needed field is achieved with the help of dipole and quadrupole magnets. But the linear motion is unstable to perturbations in the focusing fields and excludes any betatron tune spread over particle's energy, which is beneficial for the beam stability. Moreover, the focusing strength depends on the particle's energy deviation from the designed one. These undesired effects are usually suppressed by the addition of non- linear elements, such as sextupoles and octupoles. Unfortunately, these additions make focusing potential nonintegrabe, leading to the limitation of stable phase space area. Recent research showed, that there exist non-linear potentials that, nevertheless, are fully integrable [1, 2]. Thus, there is an opportunity to create stable non-linear focusing optics that provides high betatron tune spread and mitigate magnetic field errors. This approach is to be implemented in electron storage ring called Integrable Optics Test Accelerator (IOTA) at Fermilab [3]. The main goal is to show the possibility to implement the nonlinear integrable system in a realistic accelerator design. Development of a magnet, which creates exact non-linear potential, is a technically complicated task, that takes much time and efforts. However, it is possible to expand the non- linear potential into superposition of multipoles of even orders (i.e quadrupole, octupole, etc.). Though this series is not convergent in the entire space, a few low order terms can be used to model the true potential in some limited area. In turn, each of the multipole potentials can be generated by speci ed electrical current in PCB coiled on the surface of storage ring's vacuum tube. Such approach is much cheaper than the creation of a magnet of solid iron with high current coil and faster than the development of \exact" non-linear magnet, but it can help identify some features of non-linear integrable optics application. The goal of this research work is to find analytically the angular distribution of electrical current on the surface of the vacuum tube, which creates inside it multipole potential with given parameters. Hereafter, there is the task to develop a model of flexible printed circuit board (PCB), which could be coiled around the vacuum tube in order to provide the desired current distribution.

Vorobev, Viktor↗
DOE OSTI2018

Less is more: Sampling chemical space with active learning

We present the development of accurate and transferable machine learning (ML) potentials for predicting molecular energetics is a challenging task. The process of data generation to train such ML potentials is a task neither well understood nor researched in detail. In this work, we present a fully automated approach for the generation of datasets with the intent of training universal ML potentials. It is based on the concept of active learning (AL) via Query by Committee (QBC), which uses the disagreement between an ensemble of ML potentials to infer the reliability of the ensemble’s prediction. QBC allows the presented AL algorithm to automatically sample regions of chemical space where the ML potential fails to accurately predict the potential energy. AL improves the overall fitness of ANAKIN-ME (ANI) deep learning potentials in rigorous test cases by mitigating human biases in deciding what new training data to use. AL also reduces the training set size to a fraction of the data required when using naive random sampling techniques. To provide validation of our AL approach, we develop the COmprehensive Machine-learning Potential (COMP6) benchmark (publicly available on GitHub) which contains a diverse set of organic molecules. Active learning-based ANI potentials outperform the original random sampled ANI-1 potential with only 10% of the data, while the final active learning-based model vastly outperforms ANI-1 on the COMP6 benchmark after training to only 25% of the data. Finally, we show that our proposed AL technique develops a universal ANI potential (ANI-1x) that provides accurate energy and force predictions on the entire COMP6 benchmark. Finally, this universal ML potential achieves a level of accuracy on par with the best ML potentials for single molecules or materials, while remaining applicable to the general class of organic molecules composed of the elements CHNO.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2022

Local conformations and heterogeneities in structures and dynamics of isotactic polypropylene adsorbed onto carbon fiber

Carbon fiber (CF) reinforced polymers (CFRPs) have experienced widespread use in various industries. One of the most important parameters that controls the macroscopic property of CFRPs is the interface between a polymer matrix and CF. There is growing evidence to suggest the formation of a bound polymer layer (BPL), i.e., polymer chains that physically adsorb on a filler surface. However, this interface is always in contact with the thicker part of a polymer matrix, rendering its understanding a difficult task. Therein, we use CF-reinforced isotactic polypropylene (iPP) as a rational CFRP. To characterize the BPL on the CF surface, we extracted it from the CFRP using solvent-rinsing with p-xylene. The physical and thermal properties of the BPL were characterized by differential scanning calorimetry and thermogravimetric analysis, while its microscopic structures and dynamics were probed by small-angle neutron scattering and quasi-elastic neutron scattering (QENS) techniques. Subsequently, we employed atomistic molecular dynamics (MD) simulations to complement the QENS results above the bulk melting temperature and reveal details that were experimentally inaccessible. In this work, we observed that the degree of crystallinity of the BPL was quite lower than the bulk, while the melting temperature of the BPL remained the same as the bulks. Within the given length and time scales probed by QENS and MD, we also observed that most of the bound chains were mobile, with the formation of a high-density region (less than 1 nm in thickness) near the CF surface. The segmental dynamics of the bound chains probed by both QENS and MD were also much faster than those of the free chains, possibly due to the presence of a free surface region at the topmost surface of the BPL. Furthermore, the MD results demonstrated that the backbone chains and side groups lie nearly flat on the CF surface, which is the driving force for the flattening process of the iPP BPL to overcome the conformational entropy loss in the total free energy.

36 MATERIALS SCIENCE↗
DOE OSTI2026

Autonomous Coupler Alignment Using Position-Based Visual Servoing in a ROS 2 Framework

As robotic arms are becoming increasingly common alongside humans as collaborative robots, their high precision in motion enables tasks to be performed at significantly higher speeds with reduced disruption in the environment. The Fermi National Accelerator Laboratory is exploring this application by incorporating a UR16e from Universal Robots in a cleanroom setting during assembly of couplers to superconducting radio frequency cavities as part of the PIP-II project. The goal of the robotic assembly process is to precisely position the UR16e robot so that the coupler flange, mounted on the robot’s end-effector, is accurately aligned with and pressed against the cavity flange, requiring only final fastening by a lab technician. This thesis builds upon an initial system in which the robotic process was limited to the alignment phase using position-based visual servoing with an eye-in-hand camera to only align the coupler to the cavity with an offset distance. The objective of this thesis is to further advance autonomous robotic assembly by extending the process. To this end, the entire software framework was reconstructed, as the previous development environment posed significant challenges in modifying the software and adapting to hardware changes. The main contributions of this thesis are as follows: (i) a modular and scalable software framework based on ROS~2 was developed to facilitate performance expansion and interchangeability of software and hardware components; (ii) the desired alignment position for position-based visual servoing was parameterized to enable flexible configuration; and (iii) a methodology was developed to close the offset distance between the coupler and cavity utilizing the internal force-torque sensing capability of the UR16e, as visual feedback is not available during the offset-closing phase. The proposed autonomous robotic assembly reduces assembly time and technician involvement, thereby minimizing the risk of airborne particulate contamination, which is essential in the cleanroom setting. Moreover, the ROS~2-based framework provides a foundation for further expansion and continued advancement of robotic automation in Fermilab.

Giffen, Nickolas [Northern Illinois U.]↗
DOE OSTI2021

Accuracy of the time-averaged ponderomotive approximation for laser-plasma accelerator modeling

Reliable modeling of laser-plasma accelerators, where a short and intense laser pulse propagates in an underdense plasma over long distances, is a computationally challenging task. This is due to the great disparity among the scales involved in the modeling, ranging from the micrometer scale of the laser wavelength to, for instance, the meter scale of the laser-plasma interaction length for a multi-GeV-class laser-plasma accelerator. To reduce such imbalance, the time-averaged ponderomotive approximation may be used, where the plasma particle dynamics is analytically averaged over the laser frequency, and only spatiotemporal scales associated with the laser envelope are retained in the calculations, resulting in significant computational savings. Here, we characterize the accuracy and robustness of the time-averaged ponderomotive approximation for a range of laser parameters of interest for present and future laser-plasma accelerators, and we show that the error introduced by the averaging process is small in all relevant cases.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2021

NRC Multiphysics Analysis Capability Deployment FY21: Part 3

This report details the progress and activities of Idaho National Laboratory (INL) on the Nuclear Regulatory Commission (NRC) project “Development and Modeling Support for Advanced Non-Light Water Reactors.” The deliverables completed for this report are: Deliverable 1c: the capability to model gas mixtures was added to Pronghorn. A test problem mimicking the conditions achieved in a depressurized loss of forced cooling (DLOFC) event was solved with both RELAP-5 and Pronghorn. Pronghorn employed a finite vol ume method with the Kurganov-Tadmor discretization. The comparison between the mass fraction spatial profiles computed with RELAP-5 and Pronghorn clearly shows the presence of numerical artifacts (i.e., overly diffusive behavior at low Mach numbers). We confirmed that the problem disappears at higher Mach numbers. We recommend future work on the implementation of a low Mach finite volume formulation to better treat low Mach number problems. Deliverable 2a: we demonstrated two approaches to model the radiation/conduction/natural convection heat transfer across a stagnant gas for the PBMR-400 design using Pronghorn. The first approach is based on the net radiation method, which relies on the computation of view factors with the Multiphysics Object-Oriented Simulation Environment (MOOSE) ray tracing capability. The second method is a traditional thermal resistance approach. The test problems include both 2D and 3D geometries. In all cases, the results show very good agreement during a DLOFC transient. This confirms that the faster thermal resistance method produces solutions that are equivalent to the net radiation method for this geometry. Deliverable 3d: we demonstrated the use of the advection kernel for the delayed neutron precursor equation in Griffin with a 2D MSFR model. The results appear physical but further verification is recommended. We also recommend the addition of conjugate heat transfer to compute the temperatures and model the thermomechanic behavior of the reflectors and other structures. Significant memory and performance issues were encountered in the 3D axisymmetric model. Future work is recommended in this area. Task 8g: this task allows multidimensional MOOSE applications to be coupled to system codes (RELAP-7 and SAM). We implemented a faster multiphysics iteration coupling algorithm, which provides an overall 6× acceleration of the 3D-1D coupling of the core multidi- mensional fluid flow solver and the 1D primary and secondary loop model.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗
DOE OSTI2025

Multi-fidelity learning for interatomic potentials: low-level forces and high-level energies are all you need

The promise of machine learning interatomic potentials (MLIPs) has led to an abundance of public quantum mechanical (QM) training datasets. The quality of an MLIP is directly limited by the accuracy of the energies and atomic forces in the training dataset. Unfortunately, most of these datasets are computed with relatively low-accuracy QM methods, e.g. density functional theory with a moderate basis set. Due to the increased computational cost of more accurate QM methods, e.g. coupled-cluster theory with a complete basis set (CBS) extrapolation, most high-accuracy datasets are much smaller and often do not contain atomic forces. The lack of high-accuracy atomic forces is quite troubling, as training with force data greatly improves the stability and quality of the MLIP compared to training to energy alone. Because most datasets are computed with a unique level of theory, traditional single-fidelity (SF) learning is not capable of leveraging the vast amounts of published QM data. In this study, we apply multi-fidelity learning (MFL) to train an MLIP to multiple QM datasets of different levels of accuracy, i.e. levels of fidelity. Specifically, we perform three test cases to demonstrate that MFL with both low-level forces and high-level energies yields an extremely accurate MLIP—far more accurate than a SF MLIP trained solely to high-level energies and almost as accurate as a SF MLIP trained directly to high-level energies and forces. Therefore, MFL greatly alleviates the need for generating large and expensive datasets containing high-accuracy atomic forces and allows for more effective training to existing high-accuracy energy-only datasets. Indeed, low-accuracy atomic forces and high-accuracy energies are all that are needed to achieve a high-accuracy MLIP with MFL.

36 MATERIALS SCIENCE↗
DOE OSTI2017

Data Synthesis and Data Assimilation at Global Change Experiments and Fluxnet Toward Improving Land Process Models

The project was conducted during the period from 7/1/2012 to 6/30/2017 with three major tasks: (1) data synthesis and development of data assimilation (DA) techniques to constrain modeled ecosystem feedback to climate change; (2) applications of DA techniques to improve process models at different scales from ecosystem to regions and the globe; and 3) improvements of modeling soil carbon (C) dynamics by land surface models. During this period, we have synthesized published data from soil incubation experiments (e.g., Chen et al., 2016; Xu et al., 2016; Feng et al., 2016), global change experiments (e.g., Li et al., 2013; Shi et al., 2015, 2016; Liang et al., 2016) and fluxnet (e.g., Niu et al., 2012., Xia et al., 2015; Li et al., 2016). These data have been organized into multiple data products and have been used to identify general mechanisms and estimate parameters for model improvement. We used the data sets that we collected and the DA techniques to improve model performance of both ecosystem models and global land models. The objectives are: 1) to improve model simulations of litter and soil carbon storage (e.g., Schädel et al., 2013; Hararuk and Luo, 2014; Hararuk et al., 2014; Liang et al., 2015); 2) to explore the effects of CO 2 , warming and precipitation on ecosystem processes (e.g., van Groenigen et al., 2014; Shi et al., 2015, 2016; Feng et al., 2017); and 3) to estimate parameters variability in different ecosystems (e.g., Li et al., 2016). We developed a traceability framework, which was based on matrix approaches and decomposed the modeled steady-state terrestrial ecosystem carbon storage capacity into four can trace the difference in ecosystem carbon storage capacity among different biomes to four traceable components: net primary productivity (NPP), baseline C residence times, environmental scalars and climate forcing (Xia et al., 2013). With this framework, we can diagnose the differences in modeled carbon storage across ecosystems, biomes, and models. This framework has been successfully implemented by several global land models, such as CABLE (Xia et al., 2013), LPJ-GUESS (Ahlström et al., 2015), CLM (Hararuk et al., 2014; Huang et al., 2017, submitted; Shi et al., 2017, submitted), and ORCHIDEE (Huang et al., 2017, unpublished). Moreover, we have identified the theoretical foundation of the determinants of transient C storage dynamics by adding another term, C storage potential, to the steady-state traceability framework (Luo et al., 2017). The theoretical foundation of transient C storage dynamics has been applied to develop a transient traceability framework to explore the traceable components of transient C storage dynamics responded to the rising CO 2 and climate change in the two contrasting ecosystem types Duke needleleaved forest and Harvard deciduous broadleaved forest (Jiang et al., 2017, in revision). Overall, with the data synthesis, data assimilation techniques, and the steady-state and transient traceability frameworks, we have greatly improved land process models for predicting responses and feedback of terrestrial C dynamics to global change. The matrix approaches has the potential to be applied in theoretical research on nitrogen and phosphorus cycle, and therefore, the coupling of carbon-nitrogen-phosphorus.

58 GEOSCIENCES↗
DOE OSTI2023

Integrate FARM with PID controllers: IES Simulation Ecosystem Control System Development

The FARM (Feasible Actuator Range Modifier) software module is a component of the RAVEN-based FORCE framework for analysis of Integrated Energy Systems (IES). FARM was designed to support the HERON software module in the solution of the optimal dispatch problem for IES units. As the result of HERON-FARM dispatch simulation, the set-point trajectories are optimized to meet constraints on both the production variables (i.e., the variables to be optimized such as the electrical power, the hydrogen production rate, etc.) and the process variables tied to the service life of equipment (e.g., steam flowrate, vessel pressure, turbine firing temperature, etc.) at a coarse time resolution (every 10 or 100 seconds) over long time horizons (several days or weeks). In case the operational constraints need to be met at finer time resolution, the computational burden of HERON-FARM would linearly increase with the sampling rate, and sub-optimal solutions might be obtained. System responses characterized by overshoots and damped oscillations temporarily violating the imposed constraints might occur during abrupt power transients. In this report, a hierarchical control system architecture for the operation of the Dynamic Energy Technology and Integration Laboratory (DETAIL) facility constructed at INL was proposed. First, the preliminary studies on the proposed control strategy for operating the facility and the designed PI controllers were reviewed. In particular, the current approach for generating the set-point trajectories was studied, and its limits were identified. To this aim, the inclusion of a Supervisory Control layer embedding a modified version of the FARM algorithm for preserving the system safe operation over both long and real-time horizons was proposed. In this way, FARM would be applied twice, i.e., the original version (“FARM Validator”) aiding the solution of the power dispatch problem, and the modified version (“FARM Supervisory” coordinating the PI controllers to address the real-time control tasks. Despite the kernel of the two modules is the same algorithm, their roles, tasks, and capabilities are quite different. A detailed description of the role of FARM at addressing low-level control tasks is provided, along with tentative operational procedures for training the models embedded into the algorithm by using the collected experimental data.

42 ENGINEERING↗
DOE OSTI2018

Future of Work with Robotics - 18599

The DOE Environmental Management mission continues with remaining tasks that are high-hazards to human workers and high-consequences to the nation, resulting in potentially high risks to project costs and schedules. This is the mission pull. At the same time, robotic technologies are hitting a turning point with increased productivity and reliability within society, supporting a technology push. Enabling this push are better computer processing following Moore's Law, breakthroughs in biometric and learning algorithms,and riding a wave of venture funding for intended commercialization of mechatronic and Internet of Thing (IOT) projects. There has been overall improvements in material technologies, algorithms, and software standardization with tools/libraries such as the Robot Operating System (ROS). The internet has made knowledge ubiquitous and the cell phone sector has pushed the boundaries of systems-on-chip technology. This current growth is a race between large global technology companies, and between countries that understand that there is a prize with increased commerce based on being a leader in the field of robotics and automation. This technological advances have placed this technology solidly within the next 4. Industrial Revolution being tracked by the World Economic Council. Policies and law can slow this progression down, but technological changes in adjacent fields will put additional pressure on human workers by being more economical and in saving lives. A fundamental issue is will this change be gradual or disruptive such as with the forecasted singularity. There is also a realization with automation that the transition of human workers to other fields may be temporary, but in the long run, workers may run out of viable work and long-term occupations. The obvious future of human jobs are most assured by those who can embrace the change through the design, development, installment, and maintenance of the technology. Younger workers can be re-educated while older humans can be re-trained. The goal of this paper is to explore all the forces that are enabling this revolution, investigating its many impacts, and speculating on the future of human jobs during nuclear remediation. Automation has made an impact to the manufacturing sector, and has since expanded to the retail and services sectors. The question is will this same automation transition, including robotics and artificial intelligence, occur to the environmental management and cleanup sector, and how should we prepare ourselves for this change. (author)

42 ENGINEERING↗
DOE OSTI2026

HydraGNN_Predictive_GFM_2026 - Ensemble of predictive graph foundation models for atomistic materials modeling

This release contains data and parameters of HydraGNN-based graph foundation models trained as a result of the work published in the pre-print "Exascale Multi-Task Graph Foundation Models for Imbalanced, Multi-Fidelity Atomistic Data" by M. Lupo Pasini et al. (https://arxiv.org/abs/2604.15380). We jointly train on 16 open first-principles datasets (544+ million structures covering 85+ elements) using a multi-task architecture with per-dataset heads and a scalable ADIOS2/DDStore data pipeline. On Frontier, we execute six large-scale DeepHyper hyperparameter optimization campaigns in FP64 and promote the top-performing message-passing models to sustained 2,048-node training, yielding a PaiNN-based lead model. The version of HydraGNN used to generate the outputs provided in this release is HydraGNN v5.0 (https://github.com/ORNL/HydraGNN/releases/tag/v5.0) The list of datasets used for the training of the graph foundation model is the following: 1) Alexandria [1] 2) ANI1x [2] 3) MPTrj [3] 4) Open Catalyst 2020 (OC20) [4] 5) Open Catalyst 2022 (OC22) [5] 6) Open Catalyst 2025 (OC25) [6] 7) Open Direct ir Capture 2023 (ODAC23) [7] 8) Open Materials 2024 (OMat24) [8] 9) Open Molecules 2025 (OMol25) [9] 10) OMol25-neutral (subset of OMol25 that contains only molecules with zero total charge) 11) OMol25-non-neutral (subset of OMol25 that contains only molecules with non-zero total charge) 12) Open Polymers 2026 (OPoly2026) [10] 13) Nabla2DFT [11] 14) QCML [12] 15) QM7X [reference 13] 16) transition1x [14] Dataset references: [1] J. Schmidt et al., “A dataset of 175k stable and metastable materials calculated with the PBEsol and SCAN functionals,” Scientific Data, vol. 9, p. 64, 2022. [2] J. S. Smith et al., “The ANI-1ccx and ANI-1x data sets, coupled-cluster and density functional theory properties for molecules,” Scientific Data, vol. 7, p. 134, 2020. [Online]. Available: https: //www.nature.com/articles/s41597-020-0473-z [3] A. Jain et al., “Commentary: The Materials Project: A materials genome approach to accelerating materials innovation,” APL Materials, vol. 1, no. 1, p. 011002, 07 2013. [Online]. Available: https://doi.org/10.1063/1.4812323 [4] L. Chanussot et al., “Open catalyst 2020 (oc20) dataset and community challenges,” ACS Catalysis, vol. 11, no. 10, pp. 6059–6072, 2021. [Online]. Available: https://doi.org/10.1021/acscatal.0c04525 [5] K. Tran et al., “Open catalyst 2022 (oc22) dataset and challenges for oxidation electrocatalysts,” ACS Catalysis, vol. 13, no. 5, pp. 3066–3084, 2023. [Online]. Available: https://doi.org/10.1021/acscatal.2c05426 [6] S. J. Sahoo et al., “The open catalyst 2025 (oc25) dataset and models for solid-liquid interfaces,” arXiv preprint arXiv:2509.17862, 2025. [Online]. Available: https://arxiv.org/abs/2509.17862 [7] A. Sriram et al., “The open DAC 2023 dataset and challenges for sorbent discovery in direct air capture,” ACS Central Science, vol. 10, no. 5, pp. 923–941, 2024. [8] L. Barroso-Luque et al., “Open materials 2024 (omat24) inorganic materials dataset and models,” 2024. [Online]. Available: https://arxiv.org/abs/2410.12771 [9] D. S. Levine et al., “The open molecules 2025 (OMol25) dataset, evaluations, and models,” 2025. [Online]. Available: https://arxiv.org/abs/2505.08762 [10] D. S. Levine et al., The open polymers 2026 (OPoly26) dataset and evaluations,” arXiv preprint arXiv:2512.23117, 2025. [Online]. Available: https://arxiv.org/abs/2512.23117 [11] K. Khrabrov et al., “Nabla2dft: A universal quantum chemistry dataset of drug-like molecules and a benchmark for neural network potentials,” in NeurIPS 2024 Datasets and Benchmarks Track, 2024. [Online]. Available: https://openreview.net/forum?id=ElUrNM9U8c [12] S. Ganscha et al., “The QCML dataset, quantum chemistry reference data from 33.5M DFT and 14.7B semi-empirical calculations,” Scientific Data, vol. 12, p. 406, 2025. [13] J. Hoja et al., “QM7-X, a comprehensive dataset of quantum-mechanical properties spanning the chemical space of small organic molecules,” Scientific Data, vol. 8, p. 43, 2021. [Online]. Available: https://www.nature.com/articles/s41597-021-00812-2 [14] M. Schreiner et al., “Transition1x - a dataset for building generalizable reactive machine learning potentials,” Scientific Data, vol. 9, p. 779, 2022. The folder "datasets_ADIOS2_format" contains the set of pre-processed datasets in Adaptable I/O System (ADIOS) format (https://www.exascaleproject.org/research-project/adios/) that have been used for the development and training of GFMs in this work. The "datasets_ADIOS2_format" directory contains 2 sub-directories, one for the version "v1" of the datasets and one for the version "v2" of the datasets. The version "v1" of the datasets provides values of the total energy as they are extracted from the original data as it was released by the respective institutions. The version "v2" of the datasets provides values of the energy that have been realigned. The realignment was performed by training a linear regression model that predicts the total energy as a function of the chemical composition of the atomistic structure, and then subtract such prediction from the original value of the total energy. Both folders "v1" and "v2" contain 16 sub-directories, each corresponding to an ADIOS2-formatted dataset The folder "DeepHyper-results" contains the configurational files and model's parameters for all the 186 HPO trials that were successfully completed by the scalable hyperparameter optimization (HPO) runs on Frontier. The content of the folder "DeepHyper-results" I structured as follows: 1) task-list.txt: list of mpnn name, jobid, and deephyper task id 2) gfm_${MPNN}_${JOBID}_0.${TASKID}: run directory with checkpoint files 3) gfm_${MPNN}: deephyper summary directory (*.csv) for each specific MPNN type 4) deephyper-experiment-${JOBID}: output and error logs for each job The file "deephyper-sorted.csv" contains the details of each HydraGNN model built and tested by HPO, obtained by merging the (*.csv) filed from each HPO run executed. Out of all the HPO trials, we selected 10 to continue the training of the respective HydraGNN models. Due to limited computational budget available in the LRN070 allocation we could not complete the training till convergence for all these 10 selected models. The folder "models" contains multiple sub-folders, one per each HydraGNN model trained. Each model sub-folder contains the parameters of each HydraGNN model, with multiple checkpoint-restarts. The list of sub-folders are as follows: 1) multidataset_hpo-BEST1-fp64 2) multidataset_hpo-BEST2-fp64 3) multidataset_hpo-BEST3-fp64 4) multidataset_hpo-BEST4-fp64 5) multidataset_hpo-BEST5-fp64 6) multidataset_hpo-BEST6-fp64 7) multidataset_hpo-BEST7-fp64 8) multidataset_hpo-BEST8-fp64 9) multidataset_hpo-BEST9-fp64 10) multidataset_hpo-BEST10-fp64 Within each one of these folders, additional auxiliary log files are provided with descriptions about how the training proceeded. The lead PaiNN-model is contained inside "multidataset_hpo-BEST6-fp64". The file "mlp_branch_weights" contains the parameters of the multi-layer perceptron (MLP) used to reconcile the predictions of the 16 output decoding heads of the HydragNN architectures. The MLP takes in input the chemical composition of the atomistic structure and predicts averaging weights to linearly mix the predictions of each output decoding head toward consolidating them into a single one. The folder "1.1billion-structure-inference" contains 1.1 billion atomistic structures randomly generated. Each structures is associated with energy and forces predicted with the lead-PaiNN model combined with the MLP model for reconciliation of the multi-branch predictions generated by the 16 output decoding heads. The folder "1.1billion-structure-inference" contains 9,300 (*.tar.gz) subdirectories, one per Frontier compute node used to execute the inference at exascale. Once uncompressed, each (*.tar.gz) subdirectory contains an ADIOS2 (*.bp) file container, where each atomistic structure is stored as a PyTorch-Geometric Data object. The file "export_dataset_environment_variables.sh" contains the environment variables that need to be set before running the HydraGNN code to reproduce the results provided in this dataset release. The code that can be used to load the ADIOS2 files, load HydraGNN models, and run inference is available at: https://github.com/ORNL/HydraGNN/releases/tag/v5.0

36 MATERIALS SCIENCE↗
DOE OSTI2024

IRIS: A Performance-Portable Framework for Cross-Platform Heterogeneous Computing

From edge to exascale, computer architectures are becoming more heterogeneous and complex. The systems typically have fat nodes, with multicore CPUs and multiple hardware accelerators such as GPUs, FPGAs, and DSPs. This complexity is causing a crisis in programming systems and performance portability. Several programming systems are working to address these challenges, but the increasing architectural diversity is forcing software stacks and applications to be specialized for each architecture. As we show, all of these approaches critically depend on their software framework for discovery, execution, scheduling, and data orchestration. To address this challenge, we believe that a more agile and proactive software framework is essential to increase performance portability and improve user productivity. To this end, we have designed and implemented IRIS: a performance-portable framework for cross-platform heterogeneous computing. IRIS can discover available resources, manage multiple diverse programming platforms (e.g., CUDA, Hexagon, HIP, Level Zero, OpenCL, OpenMP) simultaneously in the same execution, respect data dependencies, orchestrate data movement proactively, and provide for user-configurable scheduling. To simplify data movement, IRIS introduces a shared virtual device memory with relaxed consistency among different heterogeneous devices. IRIS also adds an automatic kernel workload partitioning technique using the polyhedral model so that it can resize kernels for a wide range of devices. Our evaluation on three architectures, ranging from Qualcomm Snapdragon to a Summit supercomputer node, shows that IRIS improves portability across a wide range of diverse heterogeneous architectures with negligible overhead.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2018

Machine Learning Algorithms for Matching Theories, Simulations, and Observations in Cosmology (Final Project)

The search for simple and elegant equations to explain and predict observable natural phenomena has been a driving force in scientific theory for several centuries. The great successes of Newton's laws, Maxwell's equations, the Schrodinger equation, and the theory of General Relativity have encouraged this perspective, and perhaps a beautiful equation comprising a “theory of everything” remains to be discovered. But now scientists are grappling with systems of ever-increasing complexity, fraught with nonlinearity, and with the quantities of interest often only indirectly observed. Nature may resist a simple description, and the most important discoveries of the next century may be complex theories with countless variables and parameters. The era of big data opens up a promising new approach to scientific discovery in this setting. The predictions of modern theories, even if complex or nonlinear, can be examined through detailed and computationally-intensive simulations that take days or weeks to run and fill petabytes of disk space. The upcoming generation of scientific instruments will provide petabytes of observations at unprecedented resolution and depth. The challenge is to compare data to simulations so as to test theories and identify those that best match the observations. The data are numerous and rich but are subject to noise and various systematic biases. The simulations are large and detailed but are so costly that the effective sample size is small. Moreover, the mapping from data to simulation to the parameters of interest is typically complicated and ill-conditioned. Although these challenges, and the solutions we propose, apply across the sciences, we will ground our efforts by focusing on cosmology. Here, we are truly entering the era of big data in several senses. Ever larger experiments survey larger and larger cosmological volumes, resulting in enormous amounts of pixel data to process and extract the information of cosmological interest. These include large imaging surveys, which collect all the light in broad wavelength ranges over significant fractions of the sky, and spectroscopic surveys, which measure the light as a function of wavelength to map out 3D distances. These experiments are designed to shed light on the big unanswered questions in cosmology. Distilling the petabytes of data into observed quantities for comparison with theory is a daunting task. But these large experiments do not only result in large amounts of data to analyze: if we wish to compare to theories of physics, then we need increasingly extensive predictions for the observed quantities as well, which requires ever-larger cosmological simulations that take substantial -- and at times prohibitive -- computational resources to produce. Our goal is to develop statistical and machine learning methods for using observed and simulated data to advance machine learning with applications cosmology. In particular, we focus our research efforts on the following tasks: The challenge for automated science is that it is computationally impossible and statistically dangerous to consider every possible model in order to find the best one. We will develop Bayesian Optimization based active-learning methods that accelerate both the execution of the simulations and the search for best-fitting parameters. The key idea is to make the simulations adaptive -- across resolutions, time, and parameters -- using the data to search as the simulation runs; Most ML methods operate on simple finite dimensional feature vectors. However, many cosmology and other science applications require ML methods that can operate on more complex objects as inputs or outputs such as functions, distributions, or sets and point clouds. Our goal is to develop efficient ML methods for this problem and demonstrate their applicability in Cosmology, Astrophysics, and other science problems.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2015

Many-body dispersion effects in the binding of adsorbates on metal surfaces

A correct description of electronic exchange and correlation effects for molecules in contact with extended (metal) surfaces is a challenging task for first-principles modeling. In this work, we demonstrate the importance of collective van der Waals dispersion effects beyond the pairwise approximation for organic–inorganic systems on the example of atoms, molecules, and nanostructures adsorbed on metals. We use the recently developed many-body dispersion (MBD) approach in the context of density-functional theory [Tkatchenko et al., Phys. Rev. Lett. 108, 236402 (2012) and Ambrosetti et al., J. Chem. Phys. 140, 18A508 (2014)] and assess its ability to correctly describe the binding of adsorbates on metal surfaces. We briefly review the MBD method and highlight its similarities to quantum-chemical approaches to electron correlation in a quasiparticle picture. In particular, we study the binding properties of xenon, 3,4,9,10-perylene-tetracarboxylic acid, and a graphene sheet adsorbed on the Ag(111) surface. Accounting for MBD effects, we are able to describe changes in the anisotropic polarizability tensor, improve the description of adsorbate vibrations, and correctly capture the adsorbate–surface interaction screening. Comparison to other methods and experiment reveals that inclusion of MBD effects improves adsorption energies and geometries, by reducing the overbinding typically found in pairwise additive dispersion-correction approaches.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2014

DOE-RCT-0003641 Final Technical Report

This program studied novel concepts for an Axial Flux Reluctance Machine to capture energy from marine hydrokinetic sources and compared their attributes to a Radial Flux Reluctance Machine which was designed under a prior Department of Energy program for the same application. Detailed electromagnetic and mechanical analyses were performed to determine the validity of the concept and to provide a direct comparison with the existing conventional Radial Flux Switched Reluctance Machine designed during the Advanced Wave Energy Conversion Project, DE-EE0003641. The alternate design changed the machine topology so that the flux that is switched flows axially rather than radially and the poles themselves are long radially, as opposed to the radial flux machine that has pole pieces that are long axially. It appeared possible to build an axial flux machine that should be considerably more compact than the radial machine. In an “apples to apples” comparison, the same rules with regard to generating magnetic force and the fundamental limitations of flux density hold, so that at the heart of the machine the same torque equations hold. The differences are in the mechanical configuration that limits or enhances the change of permeance with rotor position, in the amount of permeable iron required to channel the flux via the pole pieces to the air-gaps, and in the sizing and complexity of the electrical winding. Accordingly it was anticipated that the magnetic component weight would be similar but that better use of space would result in a shorter machine with accompanying reduction in housing and support structure. For the comparison the pole count was kept the same at 28 though it was also expected that the radial tapering of the slots between pole pieces would permit a higher pole count machine, enabling the generation of greater power at a given speed in some future design. The baseline Radial Flux Machine design was established during the previous DOE program. Its characteristics were tabulated for use in comparing to the Axial Flux Machine. Three basic conceptual designs for the Axial Flux Machine were considered: (1) a machine with a single coil at the inner diameter of the machine, (2) a machine with a single coil at the outside diameter of the machine, and (3) a machine with a coil around each tooth. Slight variations of these basic configurations were considered during the study. Analysis was performed on these configurations to determine the best candidate design to advance to preliminary design, based on size, weight, performance, cost and manufacturability. The configuration selected as the most promising was the multi-pole machine with a coil around each tooth. This configuration provided the least complexity with respect to the mechanical configuration and manufacturing, which would yield the highest reliability and lowest cost machine of the three options. A preliminary design was performed on this selected configuration. For this first ever axial design of the multi rotor configuration the 'apples to apples' comparison was based on using the same length of rotor pole as the axial length of rotor pole in the radial machine and making the mean radius of the rotor in the axial machine the same as the air gap radius in the radial machine. The tooth to slot ratio at the mean radius of the axial machine was the same as the tooth to slot ratio of the radial machine. The comparison between the original radial flux machine and the new axial flux machine indicates that for the same torque, the axial flux machine diameter will be 27% greater, but it will have 30% of the length, and 76% of the weight. Based on these results, it is concluded that an axial flux reluctance machine presents a viable option for large generators to be used for the capture of wave energy. In the analysis of Task 4, below, it is pointed out that our selection of dimensional similarity for the 'apples to apples' comparison did not produce an optimum axial flux design. There is torque capability to spare, implying we could reduce the magnetic structure, but the winding area, constrained by the pole separation at the inner pole radius has a higher resistance than desirable, implying we need more room for copper. The recommendation is to proceed via one cycle of optimization and review to correct this unbalance and then proceed to a detailed design phase to produce manufacturing drawings, followed by the construction of a prototype to test the performance of the machine against predicted results.

16 TIDAL AND WAVE POWER↗