Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Matrix factorization”

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 397 records · Page 22
All sourcesNASA NTRSDOE OSTI
DOE OSTI2023

Transverse momentum distributions of heavy hadrons and polarized heavy quarks

We initiate the study of transverse momentum-dependent (TMD) fragmentation functions for heavy quarks, demonstrate their factorization in terms of novel nonperturbative matrix elements in heavy-quark effective theory (HQET), and prove new TMD sum rules that arise from heavy-quark spin symmetry. We discuss the phenomenology of heavy-quark TMD FFs at B factories and find that the Collins effect, in contrast to claims in the literature, is not parametrically suppressed by the heavy-quark mass. We further calculate all TMD parton distribution functions for the production of heavy quarks from polarized gluons within the nucleon and use our results to demonstrate the potential of the future EIC to resolve TMD heavy-quark fragmentation in semi-inclusive DIS, complementing the planned EIC program to use heavy quarks as probes of gluon distributions.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2023

Computing rank‐revealing factorizations of matrices stored out‐of‐core

This paper describes efficient algorithms for computing rank-revealing factorizations of matrices that are too large to fit in main memory (RAM), and must instead be stored on slow external memory devices such as disks (out-of-core or out-of-memory). Traditional algorithms for computing rank-revealing factorizations (such as the column pivoted QR factorization and the singular value decomposition) are very communication intensive as they require many vector-vector and matrix-vector operations, which become prohibitively expensive when data is not in RAM. Randomization allows to reformulate new methods so that large contiguous blocks of the matrix are processed in bulk. The paper describes two distinct methods. The first is a blocked version of column pivoted Householder QR, organized as a “left-looking” method to minimize the number of the expensive write operations. The second method results employs a UTV factorization. It is organized as an algorithm-by-blocks to overlap computations and I/O operations. As it incorporates power iterations, it is much better at revealing the numerical rank. Numerical experiments on several computers demonstrate that the new algorithms are almost as fast when processing data stored on slow memory devices as traditional algorithms are for data stored in RAM.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2022

Fabrication, thermal analysis, and heavy ion irradiation resistance of epoxy matrix nanocomposites loaded with silane-functionalized ceria nanoparticles

This paper describes a detailed understanding of how nanofillers function as radiation barriers within the polymer matrix, and how their effectiveness is impacted by factors such as composition, size, loading, surface chemistry, and dispersion. Here, we designed a comprehensive investigation of heavy ion irradiation resistance in epoxy matrix composites loaded with surface-modified ceria nanofillers, utilizing tandem computational and experimental methods to elucidate radiolytic damage processes and relate them to chemical and structural changes observed through thermal analysis, vibrational spectroscopy, and electron microscopy. A detailed mechanistic examination supported by FTIR spectroscopy data identified the bisphenol A moiety as a primary target for degradation reactions. Results of computational modeling by the Stopping Range of Ions in Matter (SRIM) Monte Carlo simulation were in good agreement with damage analysis from surface and cross-sectional SEM imaging. All metrics indicated that ceria nanofillers reduce the damage area in polymer nanocomposites, and that nanofiller loading and homogeneity of dispersion are key to effective damage prevention. The results of this study represent a significant pathway for engineered irradiation tolerance in a diverse array of polymer nanocomposite materials. Numerous areas of materials science can benefit from utilizing this facile and effective method to extend the reliability of polymer materials.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2024

A convergent genus expansion for the plateau

We conjecture a formula for the spectral form factor of a double-scaled matrix integral in the limit of large time, large density of states, and fixed temperature. The formula has a genus expansion with a nonzero radius of convergence. To understand the origin of this series, we compare to the semiclassical theory of “encounters” in periodic orbits. In Jackiw-Teitelboim (JT) gravity, encounters correspond to portions of the moduli space integral that mutually cancel (in the orientable case) but individually grow at low energies. At genus one we show how the full moduli space integral resolves the low energy region and gives a finite nonzero answer.

2D Gravity↗
DOE OSTI2022

Stability analysis of Monte Carlo burnup calculation based on PWR fuel rods

The existing Monte Carlo burnup algorithms may give some unstable results when using multiple burnable cells during burnup calculations. A very small time step is required to maintain the stability of the system, which pose some limits for the realistic applications. In this paper, we firstly repeat the instability issue by using a Monte Carlo code based on a typical PWR fuel rod model. Then, a theoretical stability analysis was carried out by using the perturbation theory in order to get a better understanding about the property of system stability. We determine a perturbation matrix which relates some main influence factors, such as the time step and the height of fuel rod, it can help people estimate the stability of burnup calculation system and the results are quantitatively consistent with the Monte Carlo code. (authors)

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗
DOE OSTI2022

The ITensor Software Library for Tensor Network Calculations

ITensor is a system for programming tensor network calculations with an interface modeled on tensor diagrams, allowing users to focus on the connectivity of a tensor network without manually bookkeeping tensor indices. The ITensor interface rules out common programming errors and enables rapid prototyping of algorithms. After discussing the philosophy behind the ITensor approach, we show examples of each part of the interface including Index objects, the ITensor product operator, tensor factorizations, tensor storage types, algorithms for matrix product state (MPS) and matrix product operator (MPO) tensor networks, quantum number conserving block sparse tensors, and the NDTensors library. We also review publications that have used ITensor for quantum many-body physics and for other areas where tensor networks are increasingly applied. To conclude we discuss promising features and optimizations to be added in the future.

Fishman, Matthew↗
DOE OSTI2022

Codebase release 0.3 for ITensor

ITensor is a system for programming tensor network calculations with an interface modeled on tensor diagrams, allowing users to focus on the connectivity of a tensor network without manually bookkeeping tensor indices. The ITensor interface rules out common programming errors and enables rapid prototyping of algorithms. After discussing the philosophy behind the ITensor approach, we show examples of each part of the interface including Index objects, the ITensor product operator, tensor factorizations, tensor storage types, algorithms for matrix product state (MPS) and matrix product operator (MPO) tensor networks, quantum number conserving block sparse tensors, and the NDTensors library. We also review publications that have used ITensor for quantum many-body physics and for other areas where tensor networks are increasingly applied. To conclude we discuss promising features and optimizations to be added in the future.

Fishman, Matthew↗
DOE OSTI2024

Assessment of a Grade 91 Steel Forging and Seamless Pipe Section After 141,000-Hours of Operation in a Superheat Outlet Header

In this work, two unique heats of 9Cr creep strength enhanced ferritic (CSEF) steels extracted from a retired superheat outlet header after 141,000 hours of service were evaluated. These two CSEF steels were a forging manufactured to SA-182 F91 (F91) reducer and a seamless pipe produced to SA-335 P91 (P91) pipe. Their creep deformation and fracture behavior were assessed using a lever arm creep frame integrated with in-situ high-temperature digital image correlation (DIC) system. Critical metallurgical and microstructure factors, including composition, service damage, grain matrix degradation, precipitates, and inclusions were quantitatively characterized to link the performance of the two service aged F91 and P91 CSEF steels. The creep test results show the F91 and P91 steels exhibit a large variation in creep strength and creep ductility. The F91 steel fractured at 572 hours while P91 steel fractured at 1,901 hours when subjected to a test condition of 650 °C and 100 MPa. The nominal creep strains at fracture were 12.5% (F91) and 14.5% (P91), respectively. The high-resolution DIC strain measurements reveal the local creep strain in F91 was about 50% while the local creep strain in P91 was >80%. The characterization results show that the F91 steel possessed pre-existing creep damage from its time in service, a higher fraction of inclusions, and a faster matrix grain coarsening rate. These features contribute to the observed reduction in performance for the F91 steel. The context for these findings, and the importance of metallurgical risk in an integrated life management approach will be emphasized.

Wang, Yiyu↗
NASA NTRS1982

Semi-implicit transient analysis procedures for structural dynamics analysis

A semi-implicit direct time integration procedure is presented which avoids factorization of the implicit difference solution matrix. The procedure, if properly implemented, requires only vectorial calculations and hence needs the same computer core space as explicit integration procedures. Guidelines for splitting the stiffness matrix into upper and lower matrices are established, which among other things are designed to satisfy a correct transmission of rigid-body motions from element (or grid) to its adjacent elements.

Park, K. C.↗
NASA NTRS1982

An improved semi-implicit method for structural dynamics analysis

A semi-implicit algorithm is presented for direct time integration of the structural dynamics equations. The algorithm avoids the factoring of the implicit difference solution matrix and mitigates the unacceptable accuracy losses which plagued previous semi-implicit algorithms. This substantial accuracy improvement is achieved by augmenting the solution matrix with two simple diagonal matrices of the order of the integration truncation error.

Park, K. C.↗
NASA NTRS1982

Autonomous satellite navigation methods using the Global Positioning Satellite System

This investigation considers the problem of autonomous satellite navigation using the NAVSTAR Global Positioning System (GPS). The major topics covered include the design, implementation, and validation of onboard navigation filter algorithms by means of computer simulations. The primary errors that the navigation filter design must minimize are computational effects and modeling inaccuracies due to limited capability of the onboard computer. The minimization of the effect of these errors is attained by applying the sequential extended Kalman filter using a factored covariance implementation with Q-matrix or dynamical model compensations. Peformance evaluation of the navigation filter design is carried out using both the CDC Cyber 170/750 computer and the PDP-11/60 computer. The results are obtained assuming the Phase I GPS constellation, consisting of six satellites, and a Landsat-D type spacecraft as the model for the user satellite orbit.

Murata, M.↗
NASA NTRS1984

Incompressible lifting-surface aerodynamics for a rotor-stator combination

Current literature on the three dimensional flow through compressor cascades deals with a row of rotor blades in isolation. Since the distance between the rotor and stator is usually 10 to 20 percent of the blade chord, the aerodynamic interference between them has to be considered for a proper evaluation of the aerothermodynamic performance of the stage. A unified approach to the aerodynamics of the incompressible flow through a stage is presented that uses the lifting surface theory for a compressor cascade of arbitrary camber and thickness distribution. The effects of rotor stator interference are represented as a linear function of the rotor and stator flows separately. The loading distribution on the rotor and stator flows separately. The loading distribution on the rotor and stator blades and the interference factor are determined concurrently through a matrix iteration process.

Ramachandra, S. M.↗
NASA NTRS1989

Data traffic reduction schemes for Cholesky factorization on asynchronous multiprocessor systems

Communication requirements of Cholesky factorization of dense and sparse symmetric, positive definite matrices are analyzed. The communication requirement is characterized by the data traffic generated on multiprocessor systems with local and shared memory. Lower bound proofs are given to show that when the load is uniformly distributed the data traffic associated with factoring an n x n dense matrix using n to the alpha power (alpha less than or equal 2) processors is omega(n to the 2 + alpha/2 power). For n x n sparse matrices representing a square root of n x square root of n regular grid graph the data traffic is shown to be omega(n to the 1 + alpha/2 power), alpha less than or equal 1. Partitioning schemes that are variations of block assignment scheme are described and it is shown that the data traffic generated by these schemes are asymptotically optimal. The schemes allow efficient use of up to O(n to the 2nd power) processors in the dense case and up to O(n) processors in the sparse case before the total data traffic reaches the maximum value of O(n to the 3rd power) and O(n to the 3/2 power), respectively. It is shown that the block based partitioning schemes allow a better utilization of the data accessed from shared memory and thus reduce the data traffic than those based on column-wise wrap around assignment schemes.

Naik, Vijay K.↗
NASA NTRS1991

Radiative transition probabilities for all vibrational levels in the X 1Sigma(+) state of HF

Recent analyses have led to an experimentally-based potential energy curve for the ground state of HF which includes nonadiabatic corrections and which joins smoothly to the long-range potential at an accurately determined dissociation limit. Using this potential curve and a new ab initio dipole moment function, accurate radiative transition probabilities among all vibrational levels of the ground state of HF have been calculated for selected rotational quantum numbers. Comparisons of Einstein A spontaneous emission coefficients, dipole moment absorption matrix elements, and Herman-Wallis factors for absorption bands are presented.

Zemke, Warren T.↗
DOE OSTI2021

Origin and resummation of threshold logarithms in the lattice QCD calculations of PDFs

Many present lattice QCD approaches to calculate the parton distribution functions (PDFs) rely on a factorization formula or effective theory expansion of certain Euclidean matrix elements in boosted hadron states. In the quasi- and pseudo-PDF methods, the matching coefficient in the factorization or expansion formula includes large logarithms near the threshold, which arise from the subtle interplay of collinear and soft divergences of an underlying 3D momentum distribution. We use the standard prescription to resum such logarithms in the Mellin-moment space at next-to-leading logarithmic accuracy, which also accounts for the Dokshitzer-Gribov-Lipatov-Altarelli-Parisi evolution, and we show that it can suppress the PDF at large x . Unlike the deep inelastic scattering and Drell-Yan cross sections, the resummation formula is away from the Landau pole. We then apply our formulation to reanalyze the recent lattice results for the pion valence PDF, and find that within the current data sensitivity, the effect of threshold resummation is marginal for the accessible moments and the PDF at large x.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
NASA NTRS2004

Fossil Signatures Using Elemental Abundance Distributions and Bayesian Probabilistic Classification

Elemental abundances (C6, N7, O8, Na11, Mg12, Al3, P15, S16, Cl17, K19, Ca20, Ti22, Mn25, Fe26, and Ni28) were obtained for a set of terrestrial fossils and the rock matrix surrounding them. Principal Component Analysis extracted five factors accounting for the 92.5% of the data variance, i.e. information content, of the elemental abundance data. Hierarchical Cluster Analysis provided unsupervised sample classification distinguishing fossil from matrix samples on the basis of either raw abundances or PCA input that agreed strongly with visual classification. A stochastic, non-linear Artificial Neural Network produced a Bayesian probability of correct sample classification. The results provide a quantitative probabilistic methodology for discriminating terrestrial fossils from the surrounding rock matrix using chemical information. To demonstrate the applicability of these techniques to the assessment of meteoritic samples or in situ extraterrestrial exploration, we present preliminary data on samples of the Orgueil meteorite. In both systems an elemental signature produces target classification decisions remarkably consistent with morphological classification by a human expert using only structural (visual) information. We discuss the possibility of implementing a complexity analysis metric capable of automating certain image analysis and pattern recognition abilities of the human eye using low magnification optical microscopy images and discuss the extension of this technique across multiple scales.

Hoover, Richard B.↗
DOE OSTI2024

Geometry-aware training of factorized layers in tensor Tucker format

Reducing parameter redundancies in neural network architectures is crucial for achieving feasible computational and memory requirements during train and inference of large networks. Given its easy implementation and flexibility, one promising approach is layer factorization, which reshapes weight tensors into a matrix format and parameterizes it as the product of two rank-r matrices. However, this family of approaches often requires an initial full-model warm-up phase, prior knowledge of a feasible rank, and it is sensitive to parameter initialization.In this work, we introduce a novel approach to train the factors of a Tucker decomposition of the weight tensors. Our training proposal proves to be optimal in locally approximating the original unfactorized dynamics and stable for the initialization. Furthermore, the rank of each mode is dynamically updated during training.We provide a theoretical analysis of the algorithm, showing convergence, approximation and local descent guarantees. The method's performance is further illustrated through a variety of experiments, showing remarkable training compression rates and comparable or even better performance than the full baseline and alternative layer factorization strategies.

Zangrando, Emanuele [Gran Sasso Science Institute ↗