Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “neural-networks”

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 55 records · Page 3
All sourcesNASA NTRSDOE OSTI
DOE OSTI2023

Structure and dynamics of $\mathrm{Fe_{90}}$$\mathrm{Si_3}$$\mathrm{O_7}$ liquids close to Earth's liquid core conditions

Using an artificial neural-network machine learning interatomic potential, we have performed molecular dynamics simulations to study the structure and dynamics of $\mathrm{Fe_{90}}$$\mathrm{Si_3}$$\mathrm{O_7}$ liquid close to the Earth's liquid core conditions. Here, the simulation results reveal that the short-range structural order (SRO) in the $\mathrm{Fe_{90}}$$\mathrm{Si_3}$$\mathrm{O_7}$ liquid is very strong. About 80% of the atoms are arranged in crystallinelike SRO motifs. In particular, ~ $70$% of Fe-centered clusters can be classified as either hexagonal-close-packed-like or icosahedrallike SRO motifs. The SRO clusters centered on Fe, Si, or O atoms are strongly intermixed and homogenously distributed throughout the liquid. The atomic structure of the liquid and the fractions of dominant SRO clusters are not sensitive to pressure/temperature used in the simulations except that the SRO of the O-centered clusters is enhanced close to inner core pressures. The O-diffusion coefficient is about two to three times larger than the Fe and Si ions and increases more rapidly in the deeper core regions.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗
DOE OSTI2025

Learning PDFs through interpretable latent representations in Mellin space

Representing the parton distribution functions (PDFs) of the proton and other hadrons through flexible, high-fidelity parametrizations has been a long-standing goal of particle physics phenomenology. This is particularly true since the chosen parametrization methodology can play an influential role in the ultimate PDF uncertainties as extracted in QCD global analyses; these, in turn, are often determinative of the reach of experiments at the LHC and other facilities to nonstandard physics, including at large 𝑥, where parametrization effects can be significant. In this study, we explore a series of encoder-decoder machine-learning (ML) models with various neural-network topologies as efficient means of reconstructing PDFs from meaningful information stored in an interpretable latent space. Given recent effort to pioneer synergies between QCD analyses and lattice-gauge calculations, we formulate a latent representation based on the behavior of PDFs in Mellin space, i.e., their integrated moments, and test the ability of various models to decode PDFs from this information faithfully. We introduce a numerical package, PDFdecoder, which implements several encoder-decoder models to reconstruct PDFs with high fidelity and use this end-to-end tool to explore how such neural-network-based models might connect PDF parametrizations to underlying properties like their Mellin moments. We additionally dissect patterns of learned correlations between encoded Mellin moments and reconstructed PDFs that suggest opportunities for further improvements to ML-based approaches to PDF parametrizations and uncertainty quantification.

Machine learning↗
DOE OSTI2024

Complex Structure of Molten FLiBe (2 Li F – Be F 2 ) Examined by Experimental Neutron Scattering, X-Ray Scattering, and Deep-Neural-Network Based Molecular Dynamics

The use of molten salts as coolants, fuels, and tritium breeding blankets in the next generation of fission and fusion nuclear reactors benefits from furthering the characterization of the molecular structure of molten halide salts, paving the way to predictive capability of the chemical and thermophysical properties of molten salts. Due to its neutronic, chemical, and thermochemical properties, 2 Li F - Be F 2 is a candidate molten salt for several fusion- and fission-reactor designs. We performed neutron and x-ray total-scattering measurements to determine the atomic structure of liquid 2 Li F - Be F 2 . We also performed and neural-network molecular-dynamics simulations to predict the structure obtained by neutron- and x-ray-diffraction experiments. The use of machine learning provides improvements to the efficiency in predicting the structure at a longer length scales than is achievable with simulations at significantly lower computational expense while retaining near accuracy. We found that the NNMD simulations accurately predicted the Be F 4 2 − oligomer formations seen in the experimental first-structure-factor peak. Our combination of high-resolution measurements with large-scale molecular dynamics provided an avenue to explore and experimentally verify the intermediate-range ordering beyond the first-nearest neighbor that has posed too many experimental and computational challenges in previous works. With a deeper understanding of the salt structure and ion ordering, the evolution of salt chemistry over the lifetime of a reactor can be better predicted, which is crucial to the licensing and operation of advanced fission and fusion reactors that employ molten salts. To this end, this work will serve as a reference for future studies of salt structure and macroscopic properties with and without the addition of solutes. Published by the American Physical Society 2024

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗
DOE OSTI2024

Multiple Lattice Instabilities and Complex Ground State in Cs 2 Ag Bi Br 6

Metal-halide perovskites (MHPs) are attracting considerable interest for optoelectronic applications, with Cs 2 Ag Bi Br 6 one of the main contenders among lead-free systems. Cs 2 Ag Bi Br 6 crystallizes in a nominally double-perovskite structure, but exhibits a soft lattice with large atomic fluctuations characteristic of MHPs. While crucial to understand electron-phonon and phonon-phonon couplings, the spatiotemporal correlations of these fluctuations remain largely unknown. Here, we reveal these correlations using comprehensive neutron and x-ray scattering measurements on Cs 2 Ag Bi Br 6 single crystals, complemented with first-principles simulations augmented with machine-learned neural-network potentials. We report the discovery of an unexpected complex modulated ground-state structure containing several hundred atoms, arising from a soft-phonon instability of the low-temperature tetragonal phase. Further, our experiments and simulations both reveal extensive correlated two-dimensional fluctuations of Br octahedra at finite temperature, arising from soft optic phonons that are strongly broadened by anhamonicity, reflecting very shallow potential wells. These results provide new insights into the atomic structure and fluctuations in MHPs, critical to understand and control their thermal and optoelectronic properties. Published by the American Physical Society 2024

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗
DOE OSTI2026

Extraction of the Collins-Soper Kernel from a Joint Analysis of Experimental and Lattice Data

We present a first joint extraction of the Collins-Soper kernel (CSK) combining experimental and lattice QCD data in the context of an analysis of transverse-momentum-dependent distributions (TMDs). Based on a neural-network parametrization, we perform a Bayesian reweighting of an existing fit of TMDs using lattice data, as well as a joint TMD fit to lattice and experimental data. We consistently find that the inclusion of lattice information shifts the central value of the CSK by approximately 10% and reduces its uncertainty by 40%–50%, highlighting the potential of lattice inputs to improve TMD extractions.

Avkhadiev, Artur [Massachusetts Inst. of Technolog↗
DOE OSTI2023

Tencoder: tensor-product encoder-decoder architecture for predicting solutions of PDEs with variable boundary data

It is widely hoped that artificial intelligence will boost data-driven surrogate models in science and engineering. However, fundamental spatial aspects of AI surrogate models remain under-studied. We investigate the ability of neural-network surrogate models to predict solutions to PDEs under variable boundary values. We do not wish to retrain the model when the boundary values change but to make them inputs to the model and infer the solution of the PDE under those boundary conditions. Such a capability is essential to making AI-based surrogate models practically useful. While simple feedforward networks are used for one-dimensional (1D) Poisson equation, an encoder-decoder architecture with a tensor-product layer is developed for the two-dimensional Poisson equation posed on a rectangular domain. We show that it is indeed possible to infer solutions to PDEs from variable boundary data using neural networks in this relatively simple setting, and point to future directions.

Kashi, Aditya↗
DOE OSTI2023

Real-time machine-learning-driven control system of a deformable mirror for achieving aberration-free X-ray wavefronts

A neural-network machine learning model is developed to control a bimorph adaptive mirror to achieve and preserve aberration-free coherent X-ray wavefronts at synchrotron radiation and free electron laser beamlines. The controller is trained on a mirror actuator response directly measured at a beamline with a real-time single-shot wavefront sensor, which uses a coded mask and wavelet-transform analysis. The system has been successfully tested on a bimorph deformable mirror at the 28-ID IDEA beamline of the Advanced Photon Source at Argonne National Laboratory. It achieved a response time of a few seconds and maintained desired wavefront shapes (e.g., a spherical wavefront) with sub-wavelength accuracy at 20 keV of X-ray energy. This result is significantly better than what can be obtained using a linear model of the mirror’s response. The developed system has not been tailored to a specific mirror and can be applied, in principle, to different kinds of bending mechanisms and actuators.

47 OTHER INSTRUMENTATION↗
DOE OSTI2024

Prediction of laser beam spatial profiles in a high-energy laser facility by use of deep learning

We adapt the significant advances achieved recently in the field of generative artificial intelligence/machine-learning to laser performance modeling in multipass, high-energy laser systems with application to high-shot-rate facilities relevant to inertial fusion energy. Advantages of neural-network architectures include rapid prediction capability, data-driven processing, and the possibility to implement such architectures within future low-latency, low-power consumption photonic networks. Four models were investigated that differed in their generator loss functions and utilized the U-Net encoder/decoder architecture with either a reconstruction loss alone or combined with an adversarial network loss. We achieved inference times of 1.3 ms for a 256 × 256 pixel near-field beam with errors in predicted energy of the order of 1% over most of the energy range. It is shown that prediction errors are significantly reduced by ensemble averaging the models with different weight initializations. These results suggest that including the temporal dimension in such models may provide accurate, real-time spatiotemporal predictions of laser performance in high-shot-rate laser systems.

47 OTHER INSTRUMENTATION↗
DOE OSTI2025

Fiats: Functional inference and training for surrogates

Fiats provides a platform for research on the training and deployment of neural-network surrogate models for computational science. Fiats also supports exploring, advancing, and combining functional, object-oriented, and parallel programming patterns in Fortran 2023. As such, the Fiats name has dual expansions: “Functional Inference And Training for Surrogates” or “Fortran Inference And Training for Science.” Fiats inference and training procedures are pure and therefore satisfy a language constraint imposed on procedure invocations inside Fortran’s parallel loop construct: do concurrent. Furthermore, the Fiats training procedures are built around a do concurrent parallel reduction. Several compilers can automatically parallelize do concurrent on Central Processing Units (CPUs) or Graphics Processing Units (GPUs). Fiats thus aims to achieve performance portability through standard language mechanisms.

Rouson, Damian [Lawrence Berkeley National Laborat↗
DOE OSTI2024

Materials genome innovation for computational software (magics) center

Functional layered material (LM) architectures will dominate nanomaterials science in this century. We have developed theory, modeling, simulation, and software and data tools that enhance understanding and AI guide synthesis, enable characterization of complex structures, and improve capabilities in the predictive design and growth of LMs. Research at the Center has focused on: Computational synthesis and characterization: AI guided synthesis and experimental synthesis of stacked LMs with tailored properties via optimized chemical vapor deposition (CVD) growth and liquid-phase exfoliation; study defects, edges, grain boundaries, wrinkling of atomic layers and their effects on chemical, mechanical, electrical, and optical properties. Far-from-equilibrium processes: Joint experimental and simulation based probe of electronic processes with NAQMD and ultrafast X-ray free-electron laser (XFEL) and ultrafast electron diffraction (UED) facilities at Stanford. Experimentally validate NAQMD by ultrafast electron diffraction and X-ray spectroscopy studies of structural and excited state dynamics, shape fluctuations, and phonon dynamics. Scalable software: Simulation engines for desktop-to-exascale platforms using low-overhead, linear-scaling QMD algorithms; divide-conquer-recombine NAQMD with electronic excitations; extended-Lagrangian reactive molecular dynamics (RMD), machine learning (ML) based neural-network quantum molecular dynamics (NNQMD), and super-state accelerated molecular dynamics (AMD) and kinetic Monte Carlo codes; thermal and electrical transport software; and design 3D architectures of LMs with desired functionality using scalable software. Distribution of software and data, and training: Software and simulation-experimental data generated within the Center are distributed to the materials science community via Berkeley Materials Project (MP) framework. We have also organized three workshops for software distribution and training at USC (Nov. 2017, Mar. 2018) and Gaithersburg, MD (Nov. 2018) to train researchers, with the last one in focused on underrepresented groups, in collaboration with Howard University which is one of the largest HBCUs. The Center supported a total of 46 personnel and 6 undergraduate students. These include 14 faculty, 11 postdoctoral research associates, 20 graduate research assistants, and mentored 6 undergraduate students. This resulted in the publications of 63 research papers that include 46 publications on Reactive and Quantum Dynamics Simulations, 13 publications on Machine Learning for Quantum Materials, and 4 publications on Quantum Computing.

2D Materials↗
DOE OSTI2025

Developing ML/AI Methods for High-Throughput Characterization of Multiple-Sensor Streams of Tokamak Dynamics for High-Speed Control (Final Report)

This project evaluated and developed new mathematical and algorithmic techniques capable of handling (in real-time) the growing amounts of data generated by modern fusion research. While existing numerical linear algebra (NLA) methods provide the backbone to classical data analysis and algorithms, these methods fundamentally do not port to distributed architectures nor do they allow low-latency data reduction for control. Motivated by the needs for modern fusion reactors, this project explored and implemented new numerical methods to characterize plasma dynamics, respond in real-time to discharge evolution, and to process massive-scale data accurately and rapidly more fully. This project links expertise in multiple-sensor diagnostics of tokamak plasma dynamics from Columbia University’s Plasma Physics Laboratory with expertise in massive-scale data reduction and extreme data control algorithms at Columbia University’s Data Science Institute. This interdisciplinary project (i) applied machine learning methods, (ii) implemented a properly-trained neural-network for very fast processing of high-speed plasma videography, and (ii) developed the applied mathematical methods, based on randomized-NLA (rNLA) routines, for data analysis, reduction, and real-time control. The Columbia University High Beta Tokamak-Extended Pulse (HBT-EP) facility provided data to test new algorithms and partnership with Columbia University's Data Sciences Institute evaluated the broader use of new algorithms for many challenging control applications.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2025

INTEGRATE – Inverse Network Transformations for Efficient Generation of Robust Airfoil and Turbine Enhancements

The INTEGRATE (Inverse Network Transformations for Efficient Generation of Robust Airfoil and Turbine Enhancements) project developed a new inverse-design capability for the aerodynamic design of wind turbine rotors using invertible neural networks. Training data was obtained from improved turbulence and transition models for RANS and hybrid RANS/LES solvers with machine-learned physics-based data-augmented corrections and then using the resulting neural-network(s) augmented RANS model to run thousands of 2-D and 3-D CFD simulations.

17 WIND ENERGY↗
DOE OSTI2025

Data Structure Alchemy

In an increasingly more data-driven world, the project set out to uncover the first principles of data-structure design, chart the immense design space they form, and build automation that can synthesize an optimal structure, or even a whole storage engine, for any given workload, hardware platform, and cost target. Data structures are at the center of every computational system and are directly responsible for its performance. Two core technical thrusts were defined: 1) Mapping design spaces for key data-centric abstractions (filters, hash functions, storage-engine layouts, neural-network topologies, blockchain protocols, image layouts, etc.). 2) Developing search & synthesis algorithms, initially analytical cost models, later neural-guided bi-level optimisers that navigate sextillions of candidate designs in seconds and materialise the best one as ready‐to-run code. This report distills the key insights, accomplishments, and impact.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2025

Machine Learning Neutrino-Nucleus Cross Sections

Neutrino-nucleus scattering cross sections are critical theoretical inputs for long-baseline neutrino oscillation experiments. However, robust modeling of these cross sections remains challenging. For a simple but physically motivated toy model of the DUNE experiment, we demonstrate that an accurate neural-network model of the cross section -- leveraging Standard Model symmetries -- can be learned from near-detector data. We then perform a neutrino oscillation analysis with simulated far-detector events, finding that the modeled cross section achieves results consistent with what could be obtained if the true cross section were known exactly. This proof-of-principle study highlights the potential of future neutrino near-detector datasets and data-driven cross-section models.

Wagman, Michael L. [Fermilab] (ORCID:0000000176701↗
DOE OSTI2026

Machine Learning Neutrino-Nucleus Cross Sections

Neutrino-nucleus scattering cross sections are critical theoretical inputs for long-baseline neutrino oscillation experiments. However, robust modeling of these cross sections remains challenging. For a simple but physically motivated toy model of the DUNE experiment, we demonstrate that an accurate neural-network model of the cross section—leveraging only Standard-Model symmetries— can be learned from near-detector data. We perform a neutrino oscillation analysis with simulated far-detector events, finding that oscillation analysis results enabled by our data-driven cross-section model approach the theoretical limit achievable with perfect prior knowledge of the cross section. We further quantify the effects of flux shape and detector resolution uncertainties as well as systematics from cross-section mismodeling. This proof-of-principle study highlights the potential of future neutrino near-detector datasets and data-driven cross-section models.

Tame-Narvaez, Karla [Fermilab] (ORCID:000000022249↗
DOE OSTI2026

Reconstructing the Stripping History of the Sagittarius Stream with Neural Networks

The Sagittarius (Sgr) Stream is produced by the ongoing disruption of the Sgr dwarf spheroidal (dSph) galaxy and is thought to contain multiple wraps that were stripped during different pericentric passages. In this study, we introduce a neural-network–based method trained on N-body simulations to infer the stripping time of Sgr Stream stars directly from their phase-space coordinates. We combine spectroscopic data from SEGUE, APOGEE DR17, and LAMOST DR7 low-resolution spectroscopic (LRS) survey with Gaia EDR3 astrometry and distance estimates from the latest StarHorse catalog to identify high-quality Sgr Stream members. Applying our method to these stars, we measure a clear metallicity gradient with stripping time, well described by a linear relation with slope ∼0.3 dex Gyr −1 . We further predict the stripping times of globular clusters previously suggested to originate from the Sgr dSph. M 54, Terzan 7, Terzan 8, and Arp 2 exhibit stripping times consistent with being currently bound to the Sgr remnant. Pal 12, Whiting 1, and NGC 2419 are inferred to have been stripped 0.9 ± 0.1, 1.1 ± 0.2, and 2.1 ± 0.2 Gyr ago, respectively. For NGC 4147 and NGC 5634, whose membership in the Sgr system remains uncertain, our analysis suggests stripping times of 1.1 ± 0.4 and 1.1 ± 0.1 Gyr, respectively, if they are ultimately confirmed as genuine Sgr members. These results demonstrate that data-driven models of dynamical stripping histories offer a promising approach for reconstructing the formation and chemical evolution of the Sgr Stream.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2025

Reduced basis approximations of parameterized dynamical partial differential equations via neural networks

Projection-based reduced order models are effective at approximating parameter-dependent differential equations that are parametrically separable. When parametric separability is not satisfied, which occurs in both linear and nonlinear problems, projection-based methods fail to adequately reduce the computational complexity. Devising alternative reduced order models is crucial for obtaining efficient and accurate approximations to expensive high-fidelity models. In this work, we develop a timestepping procedure for dynamical parameter-dependent problems, in which a neural-network is trained to propagate the coefficients of a reduced basis expansion. This results in an online stage with a computational cost independent of the size of the underlying problem. Here, we demonstrate our method on several parabolic partial differential equations, including a problem that is not parametrically separable.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2026

Component-Level Inverse Design of Transmon Qubits Using Neural Networks

Designing a superconducting qubit to realize specific Hamiltonian parameters typically requires iterating through a time and compute-intensive forward loop in which the designer chooses a layout geometry, simulates it, extracts circuit parameters such as capacitances, and refines the geometry. We study the inverse version of this task using a neural-network workflow that maps target Hamiltonian parameters directly to component-level layout parameters, which we subsequently demonstrate on a planar transmon layout. During training, we pair the inverse model with a frozen forward surrogate model and evaluate the loss in Hamiltonian space rather than in layout-parameter space. In validation against a conventional EM solver, 97% of generated designs produce usable geometries, and the inverse-plus-surrogate pipeline reaches mean percent errors of 0.73% for qubit frequency and 1.58% for anharmonicity, comparable to or below the fabrication and simulation-to-measurement uncertainty expected for academic-process transmon devices of this type. A single pipeline query takes ~60 ms on CPU, versus ~2 min for a conventional EM capacitance extraction on the same hardware, a speedup of approximately 2,000x. Batching minimizes the AI model inference overhead, reducing the runtime to 3.1 microseconds per sample on CPU and 2.6 microseconds per sample on GPU at a batch size of 2048, resulting in speedups of 3.9 x 10^7 and 4.6 x 10^7, respectively, relative to a single conventional CPU EM extraction. Our results indicate that component-level inverse design usefully extends and complements conventional EM simulation, including for small datasets on the order of 1,000 samples.

Seidel, Olivia [Fermilab; Texas U., Arlington]↗