Search NASA⌕ Search

SEARCH · Search NASA

Results for “Network structure”

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 1,783 records · Page 99

Geometric GNNs for charged particle tracking at GlueX

Nuclear physics experiments are aimed at uncovering the fundamental building blocks of matter. The experiments involve high-energy collisions that produce complex events with many particle trajectories. Tracking charged particles resulting from collisions in the presence of a strong magnetic field is critical to enable the reconstruction of particle trajectories and precise determination of interactions. It is traditionally achieved through combinatorial approaches that scale worse than linearly as the number of hits grows. Since particle hit data naturally form a point cloud and can be structured as graphs, graph neural networks (GNNs) emerge as an intuitive and effective choice for this task. In this study, we evaluate the GNN model for track finding on the data from the GlueX experiment at Jefferson Lab. We use simulation data to train the model and test on both simulation and real GlueX measurements. We demonstrate that GNN-based track finding outperforms the currently used traditional method at GlueX in terms of segment-based efficiency at a fixed purity while providing faster inferences. We show that the GNN model can achieve significant speedup by processing multiple events in batches, which exploits the parallel computation capability of graphical processing units (GPUs). Finally, we compare the GNN implementation on GPU and field-programmable gate array and describe the trade-off.

batched GNN pipeline↗

Complete characterization of the ground-space structure of two-body frustration-free Hamiltonians for qubits

The problem of finding the ground state of a frustration-free Hamiltonian carrying only two-body interactions between qubits is known to be solvable in polynomial time. It is also shown recently that, for any such Hamiltonian, there is always a ground state that is a product of single- or two-qubit states. However, it remains unclear whether the whole ground space is of any succinct structure. Here, we give a complete characterization of the ground space of any two-body frustration-free Hamiltonian of qubits. Namely, it is a span of tree tensor network states of the same tree structure. This characterization allows us to show that the problem of determining the ground-state degeneracy is as hard as, but no harder than, its classical analog.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Topologically derived dislocation theory for twist and stretch moiré superlattices in bilayer graphene

In this work, we develop a continuum dislocation description of twist and stretch moiré superlattices in two-dimensional material bilayers. The continuum formulation is based on the topological constraints introduced by the periodic dislocation network associated with the moiré structure. The approach is based on solving analytically for the structural distortion and displacement fields that satisfy the topological constraints and that minimize the total energy. The total energy is described by both the strain energy of each individual distorted layer and a Peierls-Nabarro-like interfacial contribution arising from stacking disregistry. The dislocation core emerges naturally within the formalism as a result of the competition between the two contributions. The approach presented here captures the structure and energetics of twist and stretch moiré superlattices of dislocations with arbitrary direction and character, without assuming an analytical solution a priori and while accounting naturally for dislocation-dislocation image interactions. In comparisons to atomistic simulations using classical potentials, the maximum structure deviation is 6%, while the maximum line energy deviation is 0.019 eV/Å. Several applications of our model are shown, including predicting the variation of structure with twist angle and describing dislocation line tension and junction energies.

36 MATERIALS SCIENCE↗

Machine Learning Out-of-Equilibrium Phases of Matter

Neural-network-based machine learning is emerging as a powerful tool for obtaining phase diagrams when traditional regression schemes using local equilibrium order parameters are not available, as in many-body localized (MBL) or topological phases. Nevertheless, instances of machine learning offering new insights have been rare up to now. Here we show that a single feed-forward neural network can decode the defining structures of two distinct MBL phases and a thermalizing phase, using entanglement spectra obtained from individual eigenstates. For this, we introduce a simplicial geometry-based method for extracting multipartite phase boundaries. We find that this method outperforms conventional metrics for identifying MBL phase transitions, revealing a sharper phase boundary and shedding new insight on the topology of the phase diagram. Furthermore, the phase diagram we acquire from a single disorder configuration confirms that the machine-learning-based approach we establish here can enable speedy exploration of large phase spaces that can assist with the discovery of new MBL phases. To our knowledge, this Letter represents the first example of a standard machine learning approach revealing new information on phase transitions.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

Shear Thickening and Jamming of Dense Suspensions: The "Roll" of Friction.

Particle-based simulations of discontinuous shear thickening (DST) and shear jamming (SJ) suspensions are used to study the role of stress-activated constraints, with an emphasis on resistance to gearlike rolling. Rolling friction decreases the volume fraction required for DST and SJ, in quantitative agreement with real-life suspensions with adhesive surface chemistries and "rough" particle shapes. It sets a distinct structure of the frictional force network compared to only sliding friction, and from a dynamical perspective leads to an increase in the velocity correlation length, in part responsible for the increased viscosity. The physics of rolling friction is thus a key element in achieving a comprehensive understanding of strongly shear-thickening materials.

Singh, Abhinendra↗

SymProp: Scaling Sparse Symmetric Tucker Decomposition via Symmetry Propagation

Sparse symmetric tensors are an important class of tensors, and their decompositions serve as powerful tools for revealing low-rank structures. This paper introduces SymProp, a novel approach for scaling sparse symmetric Tucker decomposition by propagating symmetry through intermediate computations. SymProp optimizes two key computational kernels: Sparse Symmetric Tensor Times Same Matrix chain (S3 TTMc) for Higher-Order Orthogonal Iteration (HOOI) and Sparse Symmetric Tensor Times Same Matrix chain Times Core (S3 TTMcTC) for Higher-Order QR Iteration (HOQRI). Our method employs a metaprogramming-based index iteration approach to efficiently handle the upper triangular parts of intermediate dense symmetric tensors. SymProp achieves up to 50.9× speedup over SPLATT and up to 360.8× over Compressed Sparse Symmetric (CSS) format on the S3 TTMc operation. Moreover, our S3 TTMc and S3 TTMcTC implementations support tensor orders four levels higher than state-of-the-art methods. Our HOQRI demonstrates superior scalability and up to a 33.6× speedup over optimized HOOI. By enabling more scalable Tucker decompositions for higher orders, decomposition ranks, and dimension sizes, SymProp opens new possibilities for analyzing complex hypergraph structures in fields such as network science, data mining, and machine learning.

Li, Zecheng [North Carolina State University]↗

Improved Evaluation of Large Network Matrices for Linear Power Flow Within Optimization Problems

This work presents methods for evaluating the Power Transfer Distribution Factor (PTDF) and Line Outage Distribution Factor (LODF) matrices by employing sparse linear algebra for large-scale computing applications. These matrices play a critical role in many power system applications, such as the Unit Commitment Problem (UC), pre- and post-contingency power flow analysis, and transmission expansion. These matrices are typically dense, which means they require a significant amount of time and memory to be computed for large networks. However, by analyzing the structure of the matrices and their computation method, it is possible to use reduced memory methods based on sparse matrix operations. This paper shows that sparse linear algebra algorithms are faster and require less memory and time than traditional dense approaches. Additionally, we explore the effect of matrix sparsification by eliminating trailing digits on power flow calculations.

large scale↗

Multiscale Data-Driven Seismic Full-Waveform Inversion With Field Data Study

Seismic full-waveform inversion (FWI), which uses iterative methods to estimate high-resolution subsurface models from seismograms, is a powerful imaging technique in exploration geophysics. In recent years, the computational cost of FWI has grown exponentially due to the increasing size and resolution of seismic data. Moreover, it is a nonconvex problem and can encounter local minima due to the limited accuracy of the initial velocity models or the absence of low frequencies in the measurements. To overcome these computational issues, we develop a multiscale data-driven FWI method based on fully convolutional networks (FCNs). In preparing the training data, we first develop a real-time style transform method to create a large set of synthetic subsurface velocity models from natural images. We then develop two convolutional neural networks with encoder-decoder structures to reconstruct the low- and high-frequency components of the subsurface velocity models, separately. To validate the performance of our data-driven inversion method and the effectiveness of the synthesized training set, we compare it with conventional physics-based waveform inversion approaches using both synthetic and field data. Finally, these numerical results demonstrate that, once our model is fully trained, it can significantly reduce the computation time and yield more accurate subsurface velocity models in comparison with conventional FWI.

58 GEOSCIENCES↗

3-D Unsteady Model for Be-Steam Reaction in Water-Cooled Ceramic Breeder Blanket

Here, the design of the water-cooled ceramic breeder (WCCB) blanket includes beryllium multiplier layers. At elevated temperature, beryllium reacts with steam in an exothermic reaction producing beryllium oxide and hydrogen. Such situation may occur in WCCB in the case of the rupture of one of the cooling pipes in the blanket module. This process occurs locally in a complex 3-D geometry of the blanket containing several different granular levels and a network of cooling pipes and structural supports. The process is also inherently unsteady since reaction rate depends on concentration of steam and pure beryllium which changes in time. In order to perform the detailed analysis of the process, the model of the reacting flowthrough porous media was developed and introduced into 3-D computational fluid dynamics code. In this model, granular beds are introduced as porous solids simplifying the model geometry and reducing typical mesh size to manageable amount of tens of millions of elements. A reaction rate between solid beryllium and steam is obtained from experimental results, and depends on temperature and concentration of the reactants. The differential equation for beryllium oxide fraction is introduced, allowing obtaining distributions of beryllium oxide in space and time. Multicomponent flow consisting of a homogenous mixture of steam and hydrogen is considered flowing through the porous solid with variable properties. Sink and source terms for steam and hydrogen fractions are determined by local beryllium oxide mass fraction source according to the molar ratios of beryllium steam reaction. The conjugated heat transfer approach is applied to calculate heat transfer in support structures as well as coolant flow, simultaneously with the porous medium steam flow in a blanket's granular beds. The model is validated using experimental data on beryllium steam reaction for granular bed samples.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Screening Compounds for Fast Pyrolysis and Catalytic Biofuel Upgrading Using Artificial Neural Networks

Abstract There is significant interest among researchers in finding economically sustainable alternatives to fossil-derived drop-in fuels and fuel additives. Fast pyrolysis, a method for converting biomass into liquid hydrocarbons with the potential for use as fuels or fuel additives, is a promising technology that can be two to three times less expensive at scale when compared to alternative approaches such as gasification and fermentation. However, many bio-oils directly derived from fast pyrolysis have a high oxygen content and high acidity, indicating poor performance in diesel engines when used as fuels or fuel additives. Thus, a combination of selective fast pyrolysis and chemical catalysis could produce tuned bioblendstocks that perform optimally in diesel engines. The variance in performance for derived compounds introduces a feedback loop in researching acceptable fuels and fuel additives, as various combustion properties for these compounds must be determined after pyrolysis and catalytic upgrading occurs. The present work aims to reduce this feedback loop by utilizing artificial neural networks trained with quantitative structure-property relationship values to preemptively screen pure component compounds that will be produced from fast pyrolysis and catalytic upgrading. The quantitative structure-property relationship values selected as inputs for models are discussed, the cetane number and sooting propensity of compounds derived from the catalytic upgrading of phenol are predicted, and the viability of these compounds as fuels and fuel additives is analyzed. The model constructed to predict cetane number has a test set prediction root-mean-squared error of 9.874 cetane units, and the model constructed to predict yield sooting index has a test set prediction root-mean-squared error of 13.478 yield sooting index units (on the unified scale).

09 BIOMASS FUELS↗

Local plate/rod descriptors of 3D trabecular bone micro-CT images from medial axis topologic analysis

Purpose: Trabecular bone microarchitecture is made of a complex network of plate and rod structures evolving with age and disease. The purpose of this article is to propose a new 3D local analysis method for the quantitative assessment of parameters related to the geometry of trabecular bone microarchitecture. Methods: The method is based on the topologic classification of the medial axis of the 3D image into branches, rods, and plates. Thanks to the reversibility of the medial axis, the classification is next extended to the whole 3D image. Finally, the percentages of rods and plates as well as their mean thicknesses are calculated. The method was applied both to simulated test images and 3D micro-CT images of human trabecular bone. Results: The classification of simulated phantoms made of plates and rods shows that the maximum error in the quantitative percentages of plate and rods is less than 6% and smaller than with the structure model index (SMI). Micro-CT images of human femoral bone taken in osteoporosis and early or advanced osteoarthritis were analyzed. Despite the large physiological variability, the present method avoids the underestimation of rods observed with other local methods. The relative percentages of rods and plates were not significantly different between osteoarthritis and osteoporotic groups, whereas their absolute percentages were in relation to an increase of rod and plate thicknesses in advanced osteoarthritis with also higher relative and absolute number of nodes. Conclusions: The proposed method is model-independent, robust to surface irregularities, and enables geometrical characterization of not only skeletal structures but entire 3D images. Its application provided more accurate results than the standard SMI on simple simulated phantoms, but the discrepancy observed on the advanced osteoarthritis group raises questions that will require further investigations. The systematic use of such a local method in the characterization of trabecular bone samples could provide new insight in bone microarchitecture changes related to bone diseases or to those induced by drugs or therapy.

60 APPLIED LIFE SCIENCES↗

Structural Modification of PECVD As{sub 50}S{sub 50} Chalcogenide-Glass Films by Femtosecond Laser Radiation

A thin film of As{sub 50}S{sub 50} chalcogenide glass modified by femtosecond laser radiation at a wavelength of 800 nm has been studied. The character of the modification has been analyzed by interferometry, atomic-force microscopy, and Raman spectroscopy. The relationship between a change in the refractive index upon the sample modification and structural changes in the glass network is determined.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

ATcT — Active Thermochemical Tables Python Interface

SF-25-140 atct is a lightweight, Python client for the ATcT v1 API that enables programmatic access to high-accuracy thermochemical data and turnkey reaction-enthalpy analysis. The package implements full v1 endpoint coverage (species lookup by ATcT ID, name, formula, SMILES, InChI, CAS RN; covariance queries; health checks) with robust error handling, retries, and environment-based configuration for local/production endpoints. Beyond data retrieval, atct provides rigorously implemented reaction calculators that propagate uncertainties via either (i) a conventional independent-errors method (0 K or 298.15 K) or (ii) covariance-aware propagation using provided covariances at 298.15 K. Typed data classes ensure transparent, reproducible data structures and carry ATcT Thermochemical Network (TN) version identifiers for provenance. Dual import paths and comprehensive examples facilitate integration into research pipelines, enabling reproducible thermochemical calculations, automated validation, and downstream method development.

Bross, DavidHamilton [Argonne National Laboratory ↗

Development and Optimization of Viable Human Platforms through 3D Printing

3D printing technology offers a unique method for creating cell cultures in a manner far more conducive to accurate representation of human tissues and systems. Here we print cellular structures capable of forming vascular networks and exhibiting qualities of natural tissues and human systems. This allows for cheaper and readily available sources for further study of biological and pharmaceutical agents.

42 ENGINEERING↗

Development of Approaches to Model Excited State Charge and Energy Transfer in Solution

The development of next generation energy conversion and catalytic systems requires fundamental understanding of the interplay between photo-excitation and the resulting proton and electron transfer processes that occur in solution. To address these challenges requires accurate and efficient methods to compute ground and excited states, as well as the ability to treat the dynamics of energy transfer in the presence of solvent fluctuations. Our team is developing accurate and efficient theoretical models for solution phase reactions. These developments provide an improved understanding of photo-initiated excitations, electron transfer, and proton coupled electron transfer processes. Our research team brings together expertise in electron and nuclear dynamics (classical and quantum), electronic structure, and analysis of solvation networks. We are developing and validating techniques for modeling condensed phase charge and energy transfer and to simulate the methods to monitor them using optical and vibrational spectroscopy.

74 ATOMIC AND MOLECULAR PHYSICS↗

GraphAlign: Graph-Enabled Machine Learning for Seismic Event Filtering

This report summarizes results from a 2 year effort to improve the current automated seismic event processing system by leveraging machine learning models that can operated over the inherent graph data structure of a seismic sensor network. Specifically, the GraphAlign project seeks to utilize prior information on which stations are more likely to detect signals originating from particular geographic regions to inform event filtering. To date, the GraphAlign team has developed a Graphical Neural Network (GNN) model to filter out false events generated by the Global Associator (GA) algorithm. The algorithm operates directly on waveform data that has been associated to an event by building a variable sized graph of station waveforms nodes with edge relations to an event location node. This builds off of previous work where random forest models were used to do the same task using hand crafted features. The GNN model performance was analyzed using an 8 week IMS/IDC dataset, and it was demonstrated that the GNN outperforms the random forest baseline. We provide additional error analysis of which events the GNN model performs well and poorly against concluded by future directions for improvements.

58 GEOSCIENCES↗

Reconstruction and Selection of Neutrino Interactions in MicroBooNE using Deep Convolutional Neural Networks

In this document, we describe a new reconstruction workflow developed for the MicroBooNE experiment. It features the use of Deep Convolutional Neural Networks trained to recognize key structures within the data sufficient for the 3D reconstruction of neutrino interactions within the detector. As a test of the reconstruction utility, the products of the reconstruction workflow are used to select inclusive charged-current (CC) $\nu_e$ and $\nu_\mu$ interactions in both simulated and real MicroBooNE data. In simulation, our $\nu_e$ and $\nu_\mu$ selections achieve an efficiency of 57% and 68\%, respectively, with a purity of 91% and 96%, respectively. We find that these selections are competitive with the inclusive selections used for the most recent MicroBooNE LEE searches. In particular, the CC-$\nu_e$ inclusive selection efficiency improves by over 20% while also improving sample purity. As a first step in quantifying potential bias, the data and Monte Carlo expectati ons are compared for both selections using the MicroBooNE open data. Within statistical and systematic uncertainties, both the electron and muon CC-inclusive event samples agree. A comparison of the real data events chosen by our work and another reconstruction framework shows that the two analyses each identify a sizeable fraction of events the other does not. This suggests that future analyses integrating the strengths of each could lead to combined gains. This work demonstrates, for the first time on real LArTPC data, state-of-the-art neutrino interaction reconstruction centered around deep learning algorithms.

43 PARTICLE ACCELERATORS↗

Quantum Gauge Networks: A New Kind of Tensor Network

Although tensor networks are powerful tools for simulating low-dimensional quantum physics, tensor network algorithms are very computationally costly in higher spatial dimensions. We introduce quantum gauge networks: a different kind of tensor network ansatz for which the computation cost of simulations does not explicitly increase for larger spatial dimensions. We take inspiration from the gauge picture of quantum dynamics, which consists of a local wavefunction for each patch of space, with neighboring patches related by unitary connections. A quantum gauge network (QGN) has a similar structure, except the Hilbert space dimensions of the local wavefunctions and connections are truncated. We describe how a QGN can be obtained from a generic wavefunction or matrix product state (MPS). All 2k-point correlation functions of any wavefunction for M many operators can be encoded exactly by a QGN with bond dimension O(M k ). In comparison, for just k = 1, an exponentially larger bond dimension of 2 M/6 is generically required for an MPS of qubits. We provide a simple QGN algorithm for approximate simulations of quantum dynamics in any spatial dimension. The approximate dynamics can achieve exact energy conservation for time-independent Hamiltonians, and spatial symmetries can also be maintained exactly. We benchmark the algorithm by simulating the quantum quench of fermionic Hamiltonians in up to three spatial dimensions.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗