Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “factorization methods”

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 91 records · Page 5
All sourcesNASA NTRSDOE OSTI
DOE OSTI2024

Human Factors Considerations in Artificial Intelligence Applications for Nuclear Power Plants

In recent years, there has been a wave of artificial intelligence (AI) technologies that offer to solve problems from shopping habits to mortgage approvals to critical systems operations. The rapidity of the development of these systems has led to both excitement and apprehension about the roles these systems should play in our modern societies. Furthermore, this paper focuses on the critical infrastructure industry, in general, and nuclear power generation, in particular, and seeks to scrutinize how we can leverage these novel technologies in human-centered ways to maintain or enhance the established high levels of reliability and resilience in these industries. First, we discuss the broader aspects of cognitive systems and activities that are critical to understanding the human-AI space. Then we explore different approaches to explainability in AI and the notions of trust. We then move on to discuss several human factors concepts and methods and how they can support the design of human-AI teams. We then explore recent research related to nuclear power that has been undertaken and evaluate the current industry and regulatory landscapes. Finally, we discuss identified research gaps and recommendations for solving these for the critical infrastructure space.

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗
DOE OSTI2025

Form-factor dependence of neutrino-nucleus cross sections using microscopic methods

To achieve its design goals, the next generation of neutrino-oscillation accelerator experiments requires percent-level predictions of neutrino-nucleus cross sections supplemented by robust estimates of the theoretical uncertainties involved. The latter arise from both approximations in solving the nuclear many-body problem and in the determination of the single- and few-nucleon quantities taken as input by many-body methods. To gauge the sensitivity of realistic nuclear many body methods to these few-nucleon quantities, we compute flux-averaged double-differential cross sections using the Green’s function Monte Carlo and spectral function methods as well as different parameterizations of the nucleon axial form factors based on either deuterium bubble-chamber data or lattice quantum chromodynamics calculations. The cross-section results are compared with available experimental data from the MiniBooNE and T2K collaborations. We also discuss the uncertainties associated with N → Δ transition form factors that enter the two-body current operator. We quantify the relations between neutrino-nucleus cross section and nucleon form factor uncertainties. These relations enable us to determine the form factor precision targets required to achieve a given cross-section precision.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2023

symPACK: A GPU-Capable Fan-Out Sparse Cholesky Solver

Sparse symmetric positive definite systems of equations are ubiquitous in scientific workloads and applications. Parallel sparse Cholesky factorization is the method of choice for solving such linear systems. Therefore, the development of parallel sparse Cholesky codes that can efficiently run on today’s large-scale heterogeneous distributed-memory platforms is of vital importance. Modern supercomputers offer nodes that contain a mix of CPUs and GPUs. To fully utilize the computing power of these nodes, scientific codes must be adapted to offload expensive computations to GPUs. We present symPACK, a GPU-capable parallel sparse Cholesky solver that uses one-sided communication primitives and remote procedure calls provided by the UPC++ library. We also utilize the UPC++ "memory kinds" feature to enable efficient communication of GPU-resident data. We show that on a number of large problems, symPACK outperforms comparable state-of-the-art GPU-capable Cholesky factorization codes by up to 14x on the NERSC Perlmutter supercomputer.

Bellavita, Julian↗
DOE OSTI2024

Machine learning for precise hit position reconstruction in Resistive Silicon Detectors

RSDs are LGAD silicon sensors with 100% fill factor, based on the principle of AC-coupled resistive read-out. Signal sharing and internal charge multiplication are the RSD key features to achieve picosecond-level time resolution and micron-level spatial resolution, thus making these sensors promising candidates as 4D-trackers for future experiments. This paper describes the use of a neural network to reconstruct the hit position of ionizing particles, an approach that can boost the performance of the RSD with respect to analytical models. The neural network has been trained in the laboratory and then validated on test beam data. The device-under-test in this work is a 450 μm-pitch matrix from the FBK RSD2 production, which achieved a resolution of about 65 μm at the DESY Test Beam Facility, a 50% improvement compared to a simple analytical reconstruction method, and a factor two better than the resolution of a standard pixel sensor of equal pitch size with binary read-out. The test beam result is compatible with the laboratory ones obtained during the neural network training, confirming the ability of the machine learning model to provide accurate predictions even in environments very different from the training one. Prospects for future improvements are also discussed.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
DOE OSTI2022

Demonstration and Evaluation of the Human-Technology Integration Function Allocation Methodology

There is an imminent need for the existing nuclear power plants to reduce their operating and maintenance (O&M) costs to remain economically viable. Digital technology, including automation, provides a significant opportunity for the existing nuclear power plant fleet to transform the way in which work is accomplished, reducing O&M costs, and allowing the fleet to remain economically competitive. One notable opportunity to significantly reduce O&M costs pertains to modifications to the plant equipment and main control room (MCR). Existing instrumentation and control (I&C) technologies in the MCR are highly analog, costly to operate and maintain, and demand a high cognitive and physical workload from plant staff (i.e., operators). Digitalizing the MCR has a range of broad economic benefits, including improved plant performance and reduced manual work. Further, digital I&C systems can fundamentally change the way in which plant staff operate the plant; this is the concept of operation. Human-technology integration is important to ensure that impacts to the concept of operation are done in a way that account for capabilities of people and technology. Human-technology integration employs human factors engineering (HFE) methods and principles to maximize the benefits of digital technology, reducing human error, improving overall decision-making and usability. The U.S. Department of Energy Light Water Reactor Sustainability Program is applying human-technology integration research to ensure digital technologies are safe, reliable, and efficient. This paper documents the demonstration of the human-technology guidance developed by the Light Water Reactor Sustainability Program from a first-of-a-kind digital I&C upgrade, specifically addressing function analysis and allocation for a new digital I&C system that included changes in automation levels. The program’s specific approach is included in this work, following lessons learned. This document serves as a resource for industry to follow in applying human-technology integration and HFE to digital modifications, specific to function analysis and allocation. The lessons learned should be considered in the planning and execution of HFE activities that support such digital modifications.

99 GENERAL AND MISCELLANEOUS↗
DOE OSTI2025

Assessing the difficulty of capturing the distribution function of neutrinos in neutron star merger simulations

The collision of two neutron stars is a rich source of information about nuclear physics. In particular, the kilonova signal following a merger can help us elucidate the role of neutron stars in nucleosynthesis, and informs us about the properties of matter above nuclear saturation. Approximate modeling of neutrinos remains an important limitation to our ability to make predictions for these observables. Part of the problem is the fermionic nature of neutrinos. By the exclusion principle, the expected value 𝑓 𝜈 for the number of neutrinos in a quantum state is at most 1. Any process producing neutrinos is suppressed by a blocking factor (1 −𝑓 𝜈 ). Recent simulations focused on neutrino physics mostly use a gray two-moment scheme to evolve neutrinos. This evolves integrals of 𝑓 𝜈 over momentum space, preventing direct calculations of blocking factors. Monte Carlo methods may be an attractive alternative, providing access to the full distribution of neutrinos. Their current implementation is, however, inadequate to estimate 𝑓 𝜈 : in our most recent simulations, a single Monte Carlo packet causes, in the worst cases, estimates of 𝑓 𝜈 to jump from 𝑓 𝜈 =0 to 𝑓 𝜈 ∼10 5 . While this is concerning, this brazen violation of the fermionic nature of neutrinos has been largely inconsequential, as the interactions used in simulations avoid direct calculations of 𝑓 𝜈 . We are, however, reaching a level of modeling at which this problem can no longer be ignored. Here, we discuss the relatively simple origin of this issue. We then show that very rough estimates of 𝑓 𝜈 can in theory be obtained in merger simulations, but that they will require a combination of unintuitive weighting schemes for Monte Carlo packets and smoothing of the neutrino distribution at coarser resolution than what the merger simulation uses.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2021

Recent Updates in ETOE-2 and MC 2 -3

The MC 2 -3 code is a multigroup cross section generation code for fast reactor applications, developed by Argonne National Laboratory (ANL) under the DOE Nuclear Energy Advanced Modeling and Simulation (NEAMS) program. In this work, first, the cross section libraries for ENDF/B-VII.1 and ENDF/B-VIII.0 data were generated using the updated ETOE-2 code. Preliminary verification and validation tests of the ENDF/B-VII.1 MC 2 -3 library were performed with the selected fast reactor benchmark problems. Processing and verifying the libraries suggested that significant efforts would be required for thoroughly verifying the ENDF/B-VII.1 MC 2 -3 library and successfully processing the ENDF/B-VIII.0 MC 2 -3 library. Secondly, the cross section generation capability of MC 2 -3 was updated with the intermediate group lattice calculation and the equivalent Dancoff-factor cell (EDC) method to significantly improve the performance of a twodimensional assembly calculation using the method of characteristics (MOC). This effort was made useful in implementing and verifying the EDC method in the Griffin cross section API. Finally, we analyzed a SFR problem, for which unusually large deviations in core eigenvalues from Monte Carlo solutions were reported, by generating multigroup cross sections with MC 2 -3, performing core calculations with DIF3D-VARIANT, and analyzing cross section and eigenvalue results against Monte Carlo solutions. The analysis confirmed that the MC 2 -3/DIF3D solutions were in good agreement with Monte Carlo solutions, providing the correct process of accurately generating broad-group cross sections with MC 2 -3

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2024

Expanding Solar Research and Generation for a Bright Energy Future (Final Technical Report)

This project supported the construction of an outdoor solar research facility for the primary purpose of characterizing panel and system lifetimes in the northern New England climate which features cold winters and warm, humid summers. The facility is designed to be integrated into future efforts to explore the impacts of the New England climate on energy storage and grid integration, as well as methods to mitigate factors like snow and dust coverage in the industrial and agricultural environment of the Intervale farming community and the McNeil wood-fired power generating station. It will provide year-round renewable energy research opportunities for faculty and students. As renewable energy generation and storage have become research priorities for UVM, for the state of Vermont, and for the United States as a whole, this facility will benefit the public on a broad range of scales.

14 SOLAR ENERGY↗
DOE OSTI2023

Searching for Novel Chemistry in Exoplanetary Atmospheres Using Machine Learning for Anomaly Detection

Abstract The next generation of telescopes will yield a substantial increase in the availability of high-quality spectroscopic data for thousands of exoplanets. The sheer volume of data and number of planets to be analyzed greatly motivate the development of new, fast, and efficient methods for flagging interesting planets for reobservation and detailed analysis. We advocate the application of machine learning (ML) techniques for anomaly (novelty) detection to exoplanet transit spectra, with the goal of identifying planets with unusual chemical composition and even searching for unknown biosignatures. We successfully demonstrate the feasibility of two popular anomaly detection methods (local outlier factor and one-class support vector machine) on a large public database of synthetic spectra. We consider several test cases, each with different levels of instrumental noise. In each case, we use receiver operating characteristic curves to quantify and compare the performance of the two ML techniques.

Astronomy & Astrophysics↗
DOE OSTI2024

Global 1 km land surface parameters for kilometer-scale Earth system modeling

Earth system models (ESMs) are progressively advancing towards the kilometer scale (“k-scale”). However, the surface parameters for land surface models (LSMs) within ESMs running at the k-scale are typically derived from coarse-resolution and outdated datasets. This study aims to develop a new set of global land surface parameters with a resolution of 1 km for multiple years from 2001 to 2020, utilizing the latest and most accurate available datasets. Specifically, the datasets consist of parameters related to land use and land cover, vegetation, soil, and topography. Differences between the newly developed 1 km land surface parameters and conventional parameters emphasize their potential for higher accuracy due to the incorporation of the most advanced and latest data sources. To demonstrate the capability of these new parameters, we conducted 1 km resolution simulations using the E3SM Land Model version 2 (ELM2) over the contiguous United States. Our results demonstrate that land surface parameters contribute to significant spatial heterogeneity in ELM2 simulations of soil moisture, latent heat, emitted longwave radiation, and absorbed shortwave radiation. On average, about 31 % to 54 % of spatial information is lost by upscaling the 1 km ELM2 simulations to a 12 km resolution. Using eXplainable Machine Learning (XML) methods, the influential factors driving the spatial variability and spatial information loss of ELM2 simulations were identified, highlighting the substantial impact of the spatial variability and information loss of various land surface parameters, as well as the mean climate conditions. The comparison against four benchmark datasets indicates that ELM generally performs well in simulating soil moisture and surface energy fluxes. The new land surface parameters are tailored to meet the emerging needs of k-scale LSM and ESM modeling with significant implications for advancing our understanding of water, carbon, and energy cycles under global change. The 1 km land surface parameters are publicly available at https://doi.org/10.5281/zenodo.10815170 (Li et al., 2024).

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2022

Exploring for Superhot Geothermal Targets in Magmatic Settings: Developing a Methodology

This paper presents preliminary results from a subset of work carried out as part of a multinational research project entitled DErisking Exploration for multiple geothermal Plays in magmatic ENvironments (DEEPEN). One objective of DEEPEN is to develop a customized approach to exploration for superhot geothermal plays in magmatic systems. This paper summarizes key geologic components, risk factors, and exploration methods for geothermal plays in magmatic settings based on a review and comparative analysis of international training sites. As part of a Play Fairway Analysis (PFA) approach to exploring for multiple play types in a single magmatic system, training data were compiled and weights assigned to various evidence layers. Two different approaches for weighting exploration datasets are described in this paper - one based on expert opinions and the other using statistical learning. Weights produced by both approaches will be input into a 3D PFA workflow that combines multiple exploration datasets to generate 3D geothermal favorability models, which will be applied to two international demonstration sites.

GEOTHERMAL ENERGY↗
DOE OSTI2025

Effective Li-Ion Transport Quantification in Composite Cathodes for All-Solid-State Batteries via Multiscale Modeling and Experiments

The tortuosity factor of composite cathodes significantly affects the rate performance of all-solid-state batteries (ASSBs) and has significant differences from systems with liquid electrolytes. Here, in this work, we report a simulation-experiment combined approach that quantifies the effective Li-ion transport in an ASSB composite cathode, which links tortuosity factor on ∼ μm scale to terminal voltage during cycling at the cell level (on ∼ cm scale). Two independent approaches of tortuosity factor quantification are considered: fitting electrochemical cycling data and verifying at different cycling rates and calculating from segmented tomography images, with the tortuosity factor quantified from both methods reaching self-consistency. The simulated terminal voltage using the quantified tortuosity factor has a small relative error of <3% compared to the experimental measurements. We find a significantly reduced value of the Bruggeman exponent of the catholyte phase (1.75), and using shape analysis, we show that rod-shaped catholyte particles play an important role in lowering the tortuosity factor.

Yao, Archie Mingze [Univ. of Michigan, Ann Arbor, ↗
DOE OSTI2026

Using Filter Methods to Guide Convergence for ADMM, with Applications to Nonnegative Matrix Factorization Problems

Nonconvex, nonlinear optimization problems arise naturally in parameter fitting and machine learning. While augmented Lagrangian methods have demonstrated robust convergence for classes of these problems, their convergence for block updates has been relatively unexplored outside of the context of the alternating direction method of multipliers (ADMM). ADMM has seen extensive use in these applications, but may exhibit uncertain convergence behavior in many practical nonconvex settings, and struggles with general nonlinear constraints. In contrast, filter methods have proved effective in enforcing convergence for sequential quadratic programming methods and interior point methods with feasibility criteria. We develop an ADMM-filter method for highly nonlinear and nonconvex problems. Here, we show convergence under mild assumptions for several types of coordinate descent schemes, and demonstrate our algorithm on nonnegative matrix factorization and completion problems in imaging and chemical spectrum analysis.

Nonconvex optimization↗
DOE OSTI2024

Physical discovery in representation learning via conditioning on prior knowledge

Recent advances in electron, scanning probe, optical, and chemical imaging and spectroscopy yield bespoke data sets containing the information of structure and functionality of complex systems. In many cases, the resulting data sets are underpinned by low-dimensional simple representations encoding the factors of variability within the data. The representation learning methods seek to discover these factors of variability, ideally further connecting them with relevant physical mechanisms. However, generally, the task of identifying the latent variables corresponding to actual physical mechanisms is extremely complex. Here, we present an empirical study of an approach based on conditioning the data on the known (continuous) physical parameters and systematically compare it with the previously introduced approach based on the invariant variational autoencoders. The conditional variational autoencoder (cVAE) approach does not rely on the existence of the invariant transforms and hence allows for much greater flexibility and applicability. Interestingly, cVAE allows for limited extrapolation outside of the original domain of the conditional variable. However, this extrapolation is limited compared to the cases when true physical mechanisms are known, and the physical factor of variability can be disentangled in full. We further show that introducing the known conditioning results in the simplification of the latent distribution if the conditioning vector is correlated with the factor of variability in the data, thus allowing us to separate relevant physical factors. We initially demonstrate this approach using 1D and 2D examples on a synthetic data set and then extend it to the analysis of experimental data on ferroelectric domain dynamics visualized via piezoresponse force microscopy.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
DOE OSTI2024

Decomposable Formulation of Transmission Constraints for Decentralized Power Systems Optimization

One of the most complicating factors in decentralized solution methods for a broad range of power system optimization problems is the modeling of power flow equations. Existing formulations for direct current power flows either have limited scalability or are very dense and unstructured, making them unsuitable for large-scale decentralized studies. Here, in this work, we present a novel sparsified variant of the injection shift factors formulation, which has a decomposable block-diagonal structure and scales well for large systems. We also propose a decentralized solution method, based on the alternating direction multiplier method, that efficiently handles transmission line outages in N-1 security requirements. Benchmarks on multizonal security-constrained unit commitment problems show that the proposed formulation and algorithm can reliably and efficiently solve interconnection-level test systems with up to 6,515 buses with no convergence or numerical issues.

Alternating-method of multipliers (ADMM)↗
DOE OSTI2024

RESOLVING THE ELECTROCHEMICAL EQUATIONS OF A SOLID OXIDE FUEL CELL FOR USE IN TRANSIENT SIMULATION AND INTEGRATION INTO CYBER-PHYSICAL SYSTEMS

A major challenge with complex cyber-physical systems stems from long model computational time that creates a mismatch between the model system and the physical system. The numerical modeling of solid oxide fuel cells (SOFCs) presents particular challenges due to the highly coupled nature of the underlying equations and the multiphysics needed to fully resolve their behavior during a transient event. To this end current approaches revolve around splitting the computational efforts into resolving temperature effects and resolving electrochemical effects. Current methods employed for the transient simulation of an SOFC for implementation in the Hybrid Performance (HyPer) facility cyberphysical plant at the National Energy Technology Laboratory reveal a distinct need for accelerated results with a high degree of stability. To this aim, an investigation into the computational time for the code reveals that the underlying electrochemical algorithm takes an order of magnitude more time than its thermal counterpart and has a tendency to vary in terms of iteration time and as such a rework of the underlying system is proposed. The primary method for accelerated electrochemical algorithm solutions is to employ higher order root finding recipes for the resolution of the highly coupled electrochemical equations. This is done with the intention to reduce the overall number of subiterations necessary for resolving voltage, current density, and species concentration, properties of the fuel cell that are all directly coupled and require nested iterative approaches. The overall objective of this approach is an order of magnitude reduction in calculation time without sacrificing stability and increasing accuracy. Specific approaches involve using both bounded and unbounded techniques, such as the False Position method and the Secant method (or if applicable Newton-Raphson) respectively, the drawbacks being slower convergence for False Position and instability for the Secant or Newton-Raphson methods. Current preliminary results on simplified versions of the parent functions involved for electrochemical calculations indicate a reduction in computational steps by a factor of two for the secant method and a factor of three for Newton-Raphson. When implemented into new modified electrochemical algorithms, the results indicate a possible order of magnitude reduction in calculation time.

Arias, Jesus↗
DOE OSTI2022

Impact of specimen preparation method on photovoltaic backsheet degradation during accelerated aging test

In this study, we highlight some important factors in the specimen preparation methods for evaluating photovoltaic backsheet properties after accelerated aging. Two different sequences are considered: Method (I) cut-then-age: cut into 1-cm wide strips and then expose to stress, and Method (II) age-then-cut: expose a larger sheet to stress and then cut into 1-cm widths for mechanical property measurements. We also compare the effect of three cutting methods, (a) tensile specimen punch, (b) paper cutter, and (c) fresh razor blade. Several commercial backsheets were evaluated, with stress exposures including (a) pressure cooker test (PCT), (b) dry ultraviolet (UV) radiation exposure, and (c) UV combined damp heat tests. Fourier transform infrared spectrometer (FTIR) and intrinsic viscosity (IV) were used to analyze the materials on unstressed materials and samples exposed to PCT. The results show that both the cutting method and the time of cutting have an impact on the backsheet mechanical properties. Additionally, under UV exposure, Method II, age-then-cut, generally resulted in a higher average value, with more variation than Method I; however, if the side strips from Method II were excluded, the variation dropped to the same level. This is because the specimens at the sides of the sheet get additional damage from UV light from the exposed sides of the sample. In contrast to UV exposure, PCT specimens prepared by Method II result in lower average values and higher variability. This is attributed to embrittlement through the bulk of the sample where the cutting of embrittled specimens appears to result in more edge defects which can then initiate a break at smaller strains. The data suggest that for UV exposures, the specimens should be cut after aging and the exposed side specimens discarded, and that for PCT exposures, the specimens should be cut before the exposure.

14 SOLAR ENERGY↗
DOE OSTI2025

Efficient lattice QCD computation of radiative-leptonic-decay form factors at multiple positive and negative photon virtualities

In previous work [D. Giusti, Methods for high-precision determinations of radiative-leptonic decay form factors using lattice QCD, Phys. Rev. D 107, 074507 (2023)], we showed that form factors for radiative leptonic decays of pseudoscalar mesons can be determined efficiently and with high precision from lattice QCD using the “three-dimensional (3D) method,” in which three-point functions are computed for all values of the current insertion time and the time integral is performed at the data-analysis stage. Here, we demonstrate another benefit of the 3D method: the form factors can be extracted for any number of nonzero photon virtualities from the same three-point functions at no extra cost. We present results for the $D_s → ℓνγ*$ vector form factor as a function of photon energy and photon virtuality, for both positive and negative virtuality, for a single ensemble with 340 MeV pion mass and 0.11 fm lattice spacing. In our analysis, we separately consider the two different time orderings and the different quark flavors in the electromagnetic current. We discuss in detail the behavior of the unwanted exponentials contributing to the three-point functions, as well as the choice of fit models and fit ranges used to remove them for various values of the virtuality. While positive photon virtuality is relevant for decays to multiple charged leptons, negative photon virtuality suppresses soft contributions and is of interest in QCD-factorization studies of the form factors.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗