Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “matrix multiplication”

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 523 records · Page 29
All sourcesNASA NTRSDOE OSTI
DOE OSTI2020

Heterogeneity effects on nondestructive assay measurements of enrichment in UF 6 cylinders

A set of measurements were performed using multiple mechanically cooled HPGe detectors at six positions around a cylinder uranium hexafluoride to assess if matrix inhomogeneity is detectible and effects the determined apparent uranium enrichment. These measurements were performed on nine 30B cylinders of uranium hexafluoride. Six of the cylinders appeared to be homogeneous and had similar apparent measured uranium enrichment, as calculated by the FRAM software, at all positions. However, three cylinders appeared to have local inhomogeneity based on the observed results. This was manifested as very low apparent enrichments, often <10% of the declared enrichment, on select positions. The side with a low apparent enrichment shows elevated daughter isotope activity (e.g., elevated 1001 keV and 766 keV peak count rate) and slightly reduced 235 U activity (lower 185.7 keV count rate). We hypothesize that these heterogenous cylinders may have experienced asymmetric solar heating, which caused volatile UF 6 to sublime preferentially away from the warmed side. A similar phenomenon was observed at a second facility where dose rate measurements confirm that the “low-enriched side” of cylinders show elevated gamma ray dose rates, likely from the removal of UF 6 attenuation and concentration of daughter-products. Care should be taken during uranium enrichment verification when applying methods that include gamma-rays associated with daughter nuclides to the determination of uranium enrichment on cylinders that are stored in sunlight and have asymmetric dose readings around the cylinder.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗
DOE OSTI2020

Distributed non-negative matrix factorization with determination of the number of latent features

The holistic analysis and understanding of the latent (that is, not directly observable) variables and patterns buried in large datasets is crucial for data-driven science, decision making and emergency response. Such exploratory analyses require devising unsupervised learning methods for data mining and extraction of the latent features, and non-negative matrix factorization (NMF) is one of the prominent such methods. NMF is based on compute-intense non-convex constrained minimization, which, for large datasets requires fast and distributed algorithms. However, current parallel implementations of NMF fail to estimate the number of latent features. In practice, identifying these features is both difficult and significant for pattern recognition and latent feature analysis, especially for large dense matrices. Here, we introduce a distributed NMF algorithm coupled with distributed custom clustering followed by a stability analysis on dense data, which we call DnMFk, to determine the number of latent variables. The results on synthetic data and the classical Swimmer data set demonstrate the accuracy of model determination while scaling nearly linearly across multiple processors for large data. Further, we employ DnMFk to determine the number of hidden features from a terabyte matrix.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2021

Toughening materials: enhancing resistance to fracture

It has been said that ‘God invented plasticity, but the Devil invented fracture!’ Both mechanisms represent the two prime modes of structural failure, respectively, plastic collapse and the rupture/breaking of a component, but the concept of developing materials with enhanced resistance to fracture can be difficult. This is because fracture resistance invariably involves a compromise—between strength and ductility, between strength and toughness—fundamentally leading to a ‘conflict’ between nano-/micro-structural damage and the mechanisms of toughening. Here, we examine the two major classes of such toughening: (i) intrinsic toughening, which occurs ahead of a crack tip and is motivated by plasticity—this is the principal mode of fracture resistance in ductile materials, and (ii) extrinsic toughening, which occurs at, or in the wake of, a crack tip and is associated with crack-tip shielding—this is generally the sole mode of fracture resistance in brittle materials. We briefly examine how these distinct mechanistic processes have been used to toughen synthetic materials—intrinsically in gradient materials and in multiple principal-element metallic alloys with the example of metallic glasses and high-entropy alloys, and extrinsically in ceramics with the example of ceramic-matrix composites—in comparison to Nature which has been especially adept in creating biological/natural materials which are toughened by one or both mechanistic classes, despite often consisting of constituents with meagre mechanical properties. The success of Nature has been driven by its ability to cultivate the development of materials with multiple length-scale hierarchical structures that display ingenious gradients and structural adaptability, a philosophy which we need to emulate and more importantly learn to synthesize to make structural materials of the future with unprecedented combinations of mechanical properties.

36 MATERIALS SCIENCE↗
NASA NTRS2003

Distributed Sensing of Carbon-Epoxy Composites and Filament Wound Pressure Vessels Using Fiber-Bragg Gratings

Multiple Fiber Bragg-gratings are embedded in carbon-epoxy laminates as well as in composite wound pressure vessel. Structural properties of such composites are investigated. The measurements include stress-strain relation in laminates and Poisson's ratio in several specimens with varying orientation of the optical fiber Bragg-sensor with respect to the carbon fiber in an epoxy matrix. Additionally, fiber Bragg gratings are bonded on the surface of these laminates and cylinders fabricated out of carbon-epoxy composites and multiple points are monitored and compared for strain measurements at several locations.

Grant, J.↗
DOE OSTI2020

Determining the ballistic threshold velocity for a composite structure using multiple failure models

A composite structure, such as a laminated composite panel, for example, comprises one or more layers or “plies” embedded in a matrix material or otherwise fixed together in an arrangement, commonly referred to as a “stack up.” Each material in the structure has a corresponding material failure model (MFM) defining the physical characteristics of that material. A ballistic threshold velocity computing device obtains the MFMs for each material in the composite structure, generates a predicted ballistic velocity threshold velocity for each MFM, and then generates a parametric model to compute a composite ballistic velocity threshold velocity for the composite structure.

36 MATERIALS SCIENCE↗
NASA NTRS2006

Multiple-Flat-Panel System Displays Multidimensional Data

The NASA Ames hyperwall is a display system designed to facilitate the visualization of sets of multivariate and multidimensional data like those generated in complex engineering and scientific computations. The hyperwall includes a 77 matrix of computer-driven flat-panel video display units, each presenting an image of 1,280 1,024 pixels. The term hyperwall reflects the fact that this system is a more capable successor to prior computer-driven multiple-flat-panel display systems known by names that include the generic term powerwall and the trade names PowerWall and Powerwall. Each of the 49 flat-panel displays is driven by a rack-mounted, dual-central-processing- unit, workstation-class personal computer equipped with a hig-hperformance graphical-display circuit card and with a hard-disk drive having a storage capacity of 100 GB. Each such computer is a slave node in a master/ slave computing/data-communication system (see Figure 1). The computer that acts as the master node is similar to the slave-node computers, except that it runs the master portion of the system software and is equipped with a keyboard and mouse for control by a human operator. The system utilizes commercially available master/slave software along with custom software that enables the human controller to interact simultaneously with any number of selected slave nodes. In a powerwall, a single rendering task is spread across multiple processors and then the multiple outputs are tiled into one seamless super-display. It must be noted that the hyperwall concept subsumes the powerwall concept in that a single scene could be rendered as a mosaic image on the hyperwall. However, the hyperwall offers a wider set of capabilities to serve a different purpose: The hyperwall concept is one of (1) simultaneously displaying multiple different but related images, and (2) providing means for composing and controlling such sets of images. In place of elaborate software or hardware crossbar switches, the hyperwall concept substitutes reliance on the human visual system for integration, synthesis, and discrimination of patterns in complex and high-dimensional data spaces represented by the multiple displayed images. The variety of multidimensional data sets that can be displayed on the hyperwall is practically unlimited. For example, Figure 2 shows a hyperwall display of surface pressures and streamlines from a computational simulation of airflow about an aerospacecraft at various Mach numbers and angles of attack. In this display, Mach numbers increase from left to right and angles of attack increase from bottom to top. That is, all images in the same column represent simulations at the same Mach number, while all images in the same row represent simulations at the same angle of attack. The same viewing transformations and the same mapping from surface pressure to colors were used in generating all the images.

Gundo, Daniel↗
NASA NTRS1994

Micromechanical Modeling of the Thermal Expansion of Graphite/copper Composites with Nonuniform Microstructure

Two micromechanical models were developed to investigate the thermal expansion of graphite/copper (Gr/Cu) composites. The models incorporate the effects of temperature-dependent material properties, matrix inelasticity, initial residual stresses due to processing history, and nonuniform fiber distribution. The first model is based on the multiple concentric cylinder geometry, with each cylinder treated as a two-phase composite with a characteristic fiber volume fractions. By altering the fiber volume fraction of the individual cylinders, unidirectional composites with radially nonuniform fiber distributions can be investigated using this model. The second model is based on the inelastic lamination theory. By varying the fiber content in the individual laminae, composites with nonuniform fiber distribution in the thickness direction can be investigated. In both models, the properties of the individual regions (cylinders or laminae) are calculated using the method of cells micromechanical model. Classical incremental plasticity theory is used to model the inelastic response of the copper matrix at the microlevel. The models were used to characterize the effects of nonuniform fiber distribution on the thermal expansion of Gr/Cu. These effects were compared to the effects of matrix plasticity, choice of stress-free temperature, and slight fiber misalignment. It was found that the radially nonuniform fiber distribution has little effect on the thermal expansion of Gr/Cu but could become significant for composites with large fiber-matrix transverse CTE and Young's modulus mismatch. The effect of nonuniform fiber distribution in the through-thickness direction of a laminate was more significant, but only approached that of the stress-free temperature for the most extreme cases that include large amounts of bending. Subsequent comparison with experimental thermal expansion data indicated the need for more accurate characterization of the graphite fiber thermomechanical properties. Correlation with cyclic data revealed the presence of a mechanism not considered in the developed models. The predicted response did, however, exhibit ratcheting behavior that has been observed experimentally in Gr/Cu. Finally, simulation of the actual fiber distribution of particular specimens had little effect on the predicted thermal expansion.

Bednarcyk, Brett A.↗
NASA NTRS2003

North Alabama Lightning Mapping Array (LMA): VHF Source Retrieval Algorithm and Error Analyses

Two approaches are used to characterize how accurately the North Alabama Lightning Mapping Array (LMA) is able to locate lightning VHF sources in space and in time. The first method uses a Monte Carlo computer simulation to estimate source retrieval errors. The simulation applies a VHF source retrieval algorithm that was recently developed at the NASA Marshall Space Flight Center (MSFC) and that is similar, but not identical to, the standard New Mexico Tech retrieval algorithm. The second method uses a purely theoretical technique (i.e., chi-squared Curvature Matrix Theory) to estimate retrieval errors. Both methods assume that the LMA system has an overall rms timing error of 50 ns, but all other possible errors (e.g., multiple sources per retrieval attempt) are neglected. The detailed spatial distributions of retrieval errors are provided. Given that the two methods are completely independent of one another, it is shown that they provide remarkably similar results. However, for many source locations, the Curvature Matrix Theory produces larger altitude error estimates than the (more realistic) Monte Carlo simulation.

Koshak, W. J.↗
DOE OSTI2025

Exploring the energy landscape of RBMs: reciprocal space insights into bosons, hierarchical learning and symmetry breaking

Deep generative models have become ubiquitous due to their ability to learn and sample from complex distributions. Despite the proliferation of various frameworks, the relationships among these models remain largely unexplored, a gap that hinders the development of a unified theory of AI learning. In this work, we address two central challenges: clarifying the connections between different deep generative models and deepening our understanding of their learning mechanisms. We focus on Restricted Boltzmann Machines (RBMs), a class of generative models known for their universal approximation capabilities for discrete distributions. By introducing a reciprocal space formulation for RBMs, we reveal a connection between these models, diffusion processes, and systems of coupled bosons. Our analysis shows that at initialization, the RBM operates at a saddle point, where the local curvature is determined by the singular values of the weight matrix, whose distribution follows the Marc̆enko-Pastur law and exhibits rotational symmetry. During training, this rotational symmetry is broken due to hierarchical learning, where different degrees of freedom progressively capture features at multiple levels of abstraction. This leads to a symmetry breaking in the energy landscape, reminiscent of Landau’s theory. This symmetry breaking in the energy landscape is characterized by the singular values and the weight matrix eigenvector matrix. We derive the corresponding free energy in a mean-field approximation. We show that in the limit of infinite size RBM, the reciprocal variables are Gaussian distributed. Our findings indicate that in this regime, there will be some modes for which the diffusion process will not converge to the Boltzmann distribution. To illustrate our results, we trained replicas of RBMs with different hidden layer sizes using the MNIST dataset. Our findings not only bridge the gap between disparate generative frameworks but also shed light on the fundamental processes underpinning learning in deep generative models.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2021

Nonlinear Matrix Approximation with Radial Basis Function Components

We introduce and investigate matrix approximation by decomposition into a sum of radial basis function (RBF) components. An RBF component is a generalization of the outer product between a pair of vectors, where an RBF function replaces the scalar multiplication between individual vector elements. Even though the RBF functions are positive definite, the summation across components is not restricted to convex combinations and allows us to compute the decomposition for any real matrix that is not necessarily symmetric or positive definite. We formulate the problem of seeking such a decomposition as an optimization problem with a nonlinear and non-convex loss function. Several modern versions of the gradient descent method, including their scalable stochastic counterparts, are used to solve this problem. We provide extensive empirical evidence of the effectiveness of the RBF decomposition and that of the gradient-based fitting algorithm. While being conceptually motivated by singular value decomposition (SVD), our proposed nonlinear counterpart outperforms SVD by drastically reducing the memory required to approximate a data matrix with the same L2 error for a wide range of matrix types. For example, it leads to 2 to 6 times memory save for Gaussian noise, graph adjacency matrices, and kernel matrices. Moreover, this proximity-based decomposition can offer additional interpretability in applications that involve, e.g., capturing the inner low-dimensional structure of the data, retaining graph connectivity structure, and preserving the acutance of images.

Rebrova, Elizaveta↗
DOE OSTI2020

Exploration of Electrochemical Reactions at Organic–Inorganic Halide Perovskite Interfaces via Machine Learning in In Situ Time-of-Flight Secondary Ion Mass Spectrometry

The instability of hybrid organic–inorganic perovskite (HOIP) devices is one of the significant challenges preventing commercialization. Exploring these phenomena is severely limited by the complexity of the intrinsic electrochemistry of HOIPs, the presence of multiple volatile and mobile ionic species, and the possible role of environmentally induced reactions at surfaces and triple-phase junctions. Here, in situ studies of the electrochemistry of methylammonium lead bromide perovskite with the Au electrode interface are reported via light- and voltage-dependent time-of-flight secondary ion mass spectrometry (ToF-SIMS) imaging of lateral perovskite heterostructures. While ToF-SIMS allows for the visualization of the chemical composition along the surface and its evolution with light and electrical bias, the interpretation of the multidimensional data obtained is often limited due to strong correlations between chemical signatures and the need to track multiple peaks at once. Here, a machine learning workflow combining the Hough transform and non-negative matrix factorization and no-negative tensor decomposition is developed to avoid this limitation and extract salient features of associated chemical changes and to separate the light- and voltage-dependent dynamics. Finally, combining these in situ characterizations and the machine learning workflow provides comprehensive information on the chemical nature of moving species, ion accumulation, and interfacial electrochemical reactions in HOIP devices.

36 MATERIALS SCIENCE↗
DOE OSTI2025

Ion Transport in Self-Assembled Peptoid Membranes with Carbon Nanotube Porin Channels

Artificial membranes that combine high ionic selectivity with mechanical robustness remain a key challenge for next-generation separation technologies. Here, we report ion transport measurements in biomimetic membranes composed of crystalline peptoid nanosheets co-assembled with carbon nanotube porins (CNTPs). Pure peptoid sheets formed defect-free, ion-impermeable membranes, which were then suspended over small SiN x nanopore apertures for ion transport measurements. Incorporation of CNTPs into the peptoid sheet matrix made these membranes ion permeable, with ion conductance values consistent with ion transport through individual and multiple carbon nanotube channels. In conclusion, the modularity and molecular order of peptoid membranes, combined with the exceptional conductance properties of CNTPs, position this platform as a versatile framework for assembling programmable, selective, and robust nanofluidic membranes that can bridge the performance gap between biological and synthetic membrane materials.

Biotechnology↗
DOE OSTI2020

Magnetic properties of γ-Fe 2 O 3 nanoparticles in a porous SiO 2 shell for drug delivery

A method is presented for synthesizing core–shell nanoparticles with a magnetic core and a porous shell suitable for drug delivery and other medical applications. The core contains multiple γ -Fe 2 O 3 nanoparticles (~15 nm) enclosed in a SiO 2 (~100–200 nm) matrix using either methyl (denoted TMOS- γ -Fe 2 O 3 ) or ethyl (TEOS- γ -Fe 2 O 3 ) template groups. Low-temperature Mössbauer spectroscopy showed that the magnetic nanoparticles have the maghemite structure, γ -Fe 2 O 3 , with all the vacancies in the octahedral sites. Saturation magnetization measurements revealed that the density of γ -Fe 2 O 3 was greater in the TMOS- γ -Fe 2 O 3 nanoparticles than TEOS- γ -Fe 2 O 3 nanoparticles, presumably because of the smaller methyl group. Magnetization measurements showed that the blocking temperature is around room temperature for the TMOS- γ -Fe 2 O 3 and around 250 K for the TEOS- γ -Fe 2 O 3 . Three dimensional topography analysis shows clearly that the magnetic nanoparticles are not only at the surface but have penetrated deep in the silica to form the core–shell structure.

Physics↗
NASA NTRS1984

A high-accuracy optical linear algebra processor for finite element applications

Optical linear processors are computationally efficient computers for solving matrix-matrix and matrix-vector oriented problems. Optical system errors limit their dynamic range to 30-40 dB, which limits their accuray to 9-12 bits. Large problems, such as the finite element problem in structural mechanics (with tens or hundreds of thousands of variables) which can exploit the speed of optical processors, require the 32 bit accuracy obtainable from digital machines. To obtain this required 32 bit accuracy with an optical processor, the data can be digitally encoded, thereby reducing the dynamic range requirements of the optical system (i.e., decreasing the effect of optical errors on the data) while providing increased accuracy. This report describes a new digitally encoded optical linear algebra processor architecture for solving finite element and banded matrix-vector problems. A linear static plate bending case study is described which quantities the processor requirements. Multiplication by digital convolution is explained, and the digitally encoded optical processor architecture is advanced.

Casasent, D.↗
NASA NTRS1987

Implementation of a block Lanczos algorithm for Eigenproblem solution of gyroscopic systems

The details of implementation of a general numerical procedure developed for the accurate and economical computation of natural frequencies and associated modes of any elastic structure rotating along an arbitrary axis are described. A block version of the Lanczos algorithm is derived for the solution that fully exploits associated matrix sparsity and employs only real numbers in all relevant computations. It is also capable of determining multiple roots and proves to be most efficient when compared to other, similar, exisiting techniques.

Gupta, Kajal K.↗
NASA NTRS1987

Implementation of a block Lanczos algorithm for eigenproblem solution of gyroscopic systems

This paper describes the details of implementation of a general numerical procedure developed for the accurate and economical computation of natural frequencies and associated modes of any elastic structure rotating along an arbitrary axis. A block version of the Lanczos algorithm is derived for the solution that fully exploits associated matrix sparsity and employs only real numbers in all relevant computations. It is also capable of determining multiple roots and proves to be most efficient when compared to other, similar, existing techniques.

Gupta, K. K.↗
NASA NTRS1992

Determining water cloud particle sizes from lidar depolarization measurements and time dependent multiple scattering coefficients

We present a technique of extracting water cloud particle size information from lidar measurements in conjunction with double scattering calculations. In this presentation, we describe the technique and give examples using data taken with the Air Force Phillips Laboratory's (Geophysics Directorate) low altitude Nd:YAG, elastic backscatter lidar. In a related presentation we describe the double scattering lidar model which we developed for this work. The technique uses simultaneous measurements of two orthogonal linear polarization components of lidar returns from water clouds or other media composed of spherical particles. Any depolarization of the incident lidar radiation backscattered by such a media can only be due to multiple scattering. The amount of depolarization is dependent on the extinction coefficient and the single scatter phase matrix, both of which are functions of position in the medium. The phase matrix is dependent on the index of refraction of the particles and the particle size distribution. Our technique is a modification of a procedure presented in Reference 1. There, particle sizes of water clouds are determined from double scattering calculations together with measurements of radiation scattered from volumes outside the lidar receiver's field of view (which can only be multiply scattered radiation). The methodology of our technique is similar but our 'probe' of the scattering phase function (and thus the particle size distribution) is different.

Garner, R. C.↗