Search NASA⌕ Search

SEARCH · Search NASA

Results for “Random Matrix”

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 937 records · Page 52

Relation of Cloud Occurrence Frequency, Overlap, and Effective Thickness Derived from CALIPSO and CloudSat Merged Cloud Vertical Profiles

A cloud frequency of occurrence matrix is generated using merged cloud vertical profile derived from Cloud-Aerosol Lidar with Orthogonal Polarization (CALIOP) and Cloud Profiling Radar (CPR). The matrix contains vertical profiles of cloud occurrence frequency as a function of the uppermost cloud top. It is shown that the cloud fraction and uppermost cloud top vertical pro les can be related by a set of equations when the correlation distance of cloud occurrence, which is interpreted as an effective cloud thickness, is introduced. The underlying assumption in establishing the above relation is that cloud overlap approaches the random overlap with increasing distance separating cloud layers and that the probability of deviating from the random overlap decreases exponentially with distance. One month of CALIPSO and CloudSat data support these assumptions. However, the correlation distance sometimes becomes large, which might be an indication of precipitation. The cloud correlation distance is equivalent to the de-correlation distance introduced by Hogan and Illingworth [2000] when cloud fractions of both layers in a two-cloud layer system are the same.

Kato, Seiji↗

Micromechanics of Composite Materials Governed by Vector Constitutive Laws

The high-fidelity generalized method of cells micromechanics theory has been extended for the prediction of the effective property tensor and the corresponding local field distributions for composites whose constituents are governed by vector constitutive laws. As shown, the shear analogy, which can predict effective transverse properties, is not valid in the general three-dimensional case. Consequently, a general derivation is presented that is applicable to both continuously and discontinuously reinforced composites with arbitrary vector constitutive laws and periodic microstructures. Results are given for thermal and electric problems, effective properties and local field distributions, ordered and random microstructures, as well as complex geometries including woven composites. Comparisons of the theory's predictions are made to test data, numerical analysis, and classical expressions from the literature. Further, classical methods cannot provide the local field distributions in the composite, and it is demonstrated that, as the percolation threshold is approached, their predictions are increasingly unreliable. XXXX It has been observed that the bonding between the fibers and matrix in composite materials can be imperfect. In the context of thermal conductivity, such imperfect interfaces have been investigated in micromechanical models by Dunn and Taya (1993), Duan and Karihaloo (2007), Nan et al. (1997) and Hashin (2001). The present HFGMC micromechanical method, derived for perfectly bonded composite materials governed by vector constitutive laws, can be easily generalized to include the effects of weak bonding between the constituents. Such generalizations, in the context of the mechanical micromechanics problem, involve introduction of a traction-separation law at the fiber/matrix interface and have been presented by Aboudi (1987), Bednarcyk and Arnold (2002), Bednarcyk et al. (2004) and Aboudi et al. (2013) and will be addressed in the future.

micromechanics↗

Mathematical efficiency calibration with uncertain source geometries using smart optimization

The In Situ Object Counting Software (ISOCS), a mathematical method developed by CANBERRA, is a well established technique for computing High Purity Germanium (HPGe) detector efficiencies for a wide variety of source shapes and sizes. In the ISOCS method, the user needs to input the geometry related parameters such as: the source dimensions, matrix composition and density, along with the source-to-detector distance. In many applications, the source dimensions, the matrix material and density may not be well known. Under such circumstances, the efficiencies may not be very accurate since the modeled source geometry may not be very representative of the measured geometry. CANBERRA developed an efficiency optimization software known as 'Advanced ISOCS' that varies the not well known parameters within user specified intervals and determines the optimal efficiency shape and magnitude based on available benchmarks in the measured spectra. The benchmarks could be results from isotopic codes such as MGAU, MGA, IGA, or FRAM, activities from multi-line nuclides, and multiple counts of the same item taken in different geometries (from the side, bottom, top etc). The efficiency optimization is carried out using either a random search based on standard probability distributions, or using numerical techniques that carry out a more directed (referred to as 'smart' in this paper) search. Measurements were carried out using representative source geometries and radionuclide distributions. The radionuclide activities were determined using the optimum efficiency and compared against the true activities. The 'Advanced ISOCS' method has many applications among which are: Safeguards, Decommissioning and Decontamination, Non-Destructive Assay systems and Nuclear reactor outages maintenance. (authors)

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Ultrafine-grained Fe-TiB 2 high-modulus nanocomposite steel with high strength and isotropic mechanical properties by laser powder bed fusion

Fe-TiB 2 metal matrix composite, also called high-modulus steels (HMSs), are of great interest for applications in fuel-efficient transportation infrastructure, aerospace, and wear industries due to their high specific stiffness and yield strength. However, conventional cast Fe-TiB 2 HMSs often contain coarse and sharp-edged TiB 2 particles which easily trigger premature cracking during loading. Here, we synthesized a Fe-TiB 2 nanocomposite HMS via laser powder bed fusion (LPBF) additive manufacturing of mixed micro-sized powders of Fe, Ti, and Fe 2 B. We investigated the microstructure formation and mechanical behavior of the Fe-TiB 2 HMS. We found that in situ chemical reaction of Ti and Fe 2 B enables the formation of TiB 2 particles at nanoscale during rapid solidification of LPBF. These nanoscale TiB 2 particles can serve as heterogeneous nucleation sites and promote the formation of ultrafine and equiaxed α-Fe grains with random crystallographic textures, which differ from many other additively manufactured (AM) metal alloys characteristic of strong crystallographic textures. As such, isotropic mechanical properties were achieved in the AM Fe-TiB 2 nanocomposite HMS with a high elastic modulus of ~ 240 GPa, an exceptional yield strength of ~ 1450 MPa, and a large plasticity of ~ 20% under compression. Quantitative analysis reveals that the high yield strength primarily originates from strengthening contributions of the ultrafine grains with an average grain size of ~450 nm, the nanoscale TiB 2 reinforcing particles of 20–180 nm, and a high density of printing-induced dislocations of the order of 10 15 m –2 . In situ synchrotron high-energy X-ray diffraction unveils the load partitioning from the softer α-Fe matrix to the stiffer and stronger TiB 2 nanoparticles, contributing to the sustained strain hardening during compression. Our work not only provides a general pathway for achieving high-performance metal matrix nanocomposites by in situ chemical reaction and precipitation of ceramic nanoparticles during additive manufacturing, but also offers mechanistic insights into the deformation mechanism of nanoparticle-reinforced HMS composites.

36 MATERIALS SCIENCE↗

Morphology of calcite crystals in clast coatings from four soils in the Mojave desert region

Pedogenic calcite-crystal coatings on clasts were examined in four soils along an altitudinal gradient on Kyle Canyon alluvium in southern Nevada. Clast coatings were studied rather than matrix carbonate to avoid the effects of soil matrix on crystallization. Six crystal sizes and shapes were recognized and distinguished. Equant micrite was the dominant crystal form with similar abundance at all elevations. The distributions of five categories of spar and microspar appear to be influenced by altitudinally induced changes in effective moisture. In the drier, lower elevation soils, crystals were equant or parallel prismatic with irregular, interlocking boundaries while in the more moist, higher elevation soils they were randomly oriented, euhedral, prismatic, and fibrous. There was little support for the supposition that Mg(+2) substitution or increased (Mg + Ca)/HCO3 ratios in the precipitating solution produced crystal elongation.

Chadwick, Oliver A.↗

Determination of single crystal thermal expansion in Uranium-6wt%Niobium shape memory alloy using in-situ diffraction and modeling of textured polycrystalline samples

In-situ neutron diffraction and ex-situ dilatometry experiments were performed on a chemically banded, quenched uranium-6wt% niobium shape memory alloy to study the impact of deformation-(detwinning-) induced texture on the evolution of thermally induced strains (and associated stresses). Thermal heating and cooling cycles between 5K and 473K were performed in-situ with neutron diffraction, and the lattice strain evolution is reported for the observed monoclinic α" phase. Comparisons between the measured diffraction strains and macroscopic dilatometry experiments reveal relationships between micro- and macro-level thermal expansion. Softening of the texture during heating suggests that twin boundary motion can accommodate the large internal thermal strains which are approximately 2% greater than that typically observed during the 175K heating interval used to age more randomly oriented polycrystalline material. Assuming weak constraint of neighboring grains in randomly textured polycrystals, the single crystal thermal expansion tensor is extracted from measurements of lattice strains over the range 5K to 473K. Predictions of polycrystalline thermal expansion, based upon this single crystal thermal expansion tensor, are shown to compare favorably with bulk thermal expansion observations of the as quenched microstructure. However, such a lower-bound estimate is insufficient to explain all aspects of the behavior of the textured material, where the matrix is not isotropic. In conclusion, it is hypothesized that relaxation processes which occur within the quenched microstructure during heating are responsible for the distinct thermal expansion behavior observed during the initial heating cycle compared to cooling and subsequent cycling.

36 MATERIALS SCIENCE↗

Modeling of Electromagnetic Scattering by Discrete and Discretely Heterogeneous Random Media by Using Numerically Exact Solutions of the Maxwell Equations

In this paper, we discuss some aspects of numerical modeling of electromagnetic scattering by discrete random medium by using numerically exact solutions of the macroscopic Maxwell equations. Typical examples of such media are clouds of interstellar dust, clouds of interplanetary dust in the Solar system, dusty atmospheres of comets, particulate planetary rings, clouds in planetary atmospheres, aerosol particles with numerous inclusions and so on. Our study is based on the results of extensive computations of different characteristics of electromagnetic scattering obtained by using the superposition T-matrix method which represents a direct computer solver of the macroscopic Maxwell equations for an arbitrary multisphere configuration. As a result, in particular, we clarify the range of applicability of the low-density theories of radiative transfer and coherent backscattering as well as of widely used effective-medium approximations.

Dlugach, Janna M.↗

Graph Metric Learning Quantifies Morphological Differences between Two Genotypes of Shoot Apical Meristem Cells in Arabidopsis

We present a method for learning “spectrally descriptive” edge weights for graphs. We generalize a previously known distance measure on graphs (Graph Diffusion Distance), thereby allowing it to be tuned to minimize an arbitrary loss function. Because all steps involved in calculating this modified GDD are differentiable, we demonstrate that it is possible for a small neural network model to learn edge weights which minimize loss. We apply this method to discriminate between graphs constructed from shoot apical meristem images of two genotypes of Arabidopsis thaliana specimens: wild-type and trm678 triple mutants with cell division phenotype. Training edge weights and kernel parameters with contrastive loss produces a learned distance metric with large margins between these graph categories. We demonstrate this by showing improved performance of a simple k-nearest-neighbors classifier on the learned distance matrix. We also demonstrate a further application of this method to biological image analysis. Once trained, we use our model to compute the distance between the biological graphs and a set of graphs output by a cell division simulator. Comparing simulated cell division graphs to biological ones allows us to identify simulation parameter regimes which characterize mutant vs. wild-type Arabidopsis cells. We find that trm678 mutant cells are characterized by increased randomness of division planes and decreased ability to avoid previous vertices between cell walls.

59 BASIC BIOLOGICAL SCIENCES↗

Tensor Decompositions for Count Data that Leverage Stochastic and Deterministic Optimization

There is growing interest to extend low-rank matrix decompositions to multi-way arrays, or tensors. One fundamental low-rank tensor decomposition is the canonical polyadic decomposition (CPD). The challenge of fitting a low-rank, nonnegative CPD model to Poisson-distributed count data is of particular interest. Several popular algorithms use local search methods to approximate the global maximum likelihood estimator from local minima. Simultaneously, a recent trend in theoretical computer science and numerical linear algebra leverages randomization to solve very large, hard problems. The typical approach is to use randomization for a fast approximation and determinism for refinement to yield effective algorithms with theoretical guarantees. Two popular algorithms for Poisson CPD reflect that emergent dichotomy: CP Alternating Poisson Regression is a deterministic algorithm and Generalized Canonical Polyadic decomposition makes use of stochastic algorithms in several variants. This work extends recent work to develop two new methods that leverage randomized and deterministic algorithms for improved accuracy and performance.

97 MATHEMATICS AND COMPUTING↗

Entanglement and confinement in coupled quantum systems

We study some general properties of coupled quantum systems. We consider simple interactions between two copies of identical Hamiltonians such as the SYK model, Pauli spin chains with random magnetic field and harmonic oscillators. Such couplings make the ground states close to the thermofield double states of the uncoupled Hamiltonians. For the coupled SYK model, we push the numerical computation further towards the thermodynamic limit so that an extrapolation in the size of the system is possible. We find good agreement between the extrapolated numerical result and the analytic result in the large-q limit. We also consider the coupled gauged matrix model and vector model, and argue that the deconfinement is associated with the loss of the entanglement, similarly to the previous observation for the coupled SYK model. The understanding of the microscopic mechanism of the confinement/deconfinement transition enables us to estimate the quantum entanglement precisely, and backs up the dual gravity interpretation which relates the deconfinement to the disappearance of the wormhole. Our results demonstrate the importance of the entanglement between the color degrees of freedom in the emergence of the bulk geometry from quantum field theory via holography.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

PILOT-SCALE FIELD VALIDATION OF THE LONG ELECTRODE ELECTRICAL RESISTIVITY TOMOGRAPHY METHOD

Field validation for the long electrode electrical resistivity tomography (LE-ERT) method was attempted in order to demonstrate the performance of the technique in imaging a simple buried target. The experiment was an approximately 1/17 scale mock-up of a region encompassing a buried nuclear waste tank on the Hanford site. The target of focus was constructed by manually forming a simulated plume within the vadose zone using a tank waste simulant. The LE-ERT results were compared to ERT using conventional point electrodes on the surface and buried within the survey domain. Using a pole-pole array, both point and long electrode imaging techniques identified the lateral extents of the pre-formed plume with reasonable fidelity, but the LE-ERT was handicapped in reconstructing the vertical boundaries. The pole-dipole and dipole-dipole arrays were also tested with the LE-ERT method and were shown to have the least favorable target properties, including the position of the reconstructed plume relative to the known plume and the intensity of false positive targets. The poor performance of the pole-dipole and dipole-dipole arrays was attributed to an inexhaustive and non-optimal coverage of data at key electrodes, as well as an increased noise for electrode combinations with high geometric factors. However, when comparing the model resolution matrix among the different acquisition strategies, the pole-dipole and dipole-dipole arrays using long electrodes were shown to have significantly higher average and maximum values than any pole-pole array. The model resolution describes how well the inversion model resolves the subsurface. Given the model resolution performance of the pole-dipole and dipole-dipole arrays, it may be worth investing in tools to understand the optimum subset of randomly distributed electrode pairs to produce maximum performance from the inversion model.

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗

Triangle Method for Dense ReLU Layers [SWR-25-72]

This software is an implementation of the methods for initializing and training neural networks to be more efficient per parameter, described more fully below and in the related publication: In theory, depth should make a ReLU network EXPONENTIALLY more efficient by enabling it to produce an exponential number of piecewise linear sections in its output. This reasoning is largely based on the work of mathematicians that have hand-constructed networks that make good use of depth. In practice however, even very deep ReLU networks that have been randomly initialized will behave identically to their shallow counterparts - missing an entire exponential dimension of efficiency. The triangle method is a first attempt at realizing the exponential potential of deep networks. Instead of randomly setting weights, we force pairs of neurons in each layer learn to build triangles (i.e. functions from [0,1] -> [0,1] that look like triangles). This is a very efficient pattern for generating lots of linear pieces because composing two triangular functions doubles the number of pieces with each composition. The triangle method is more than just a different initialization, it is a new paradigm of training. Instead of making direct updates to the matrix weights, we do an extra step of backpropagation to collect the derivatives of the loss function with respect to the shapes of the triangles, training them to tilt left or right. This process essentially holds the networks hand throughout the loss landscape and forces it to always use depth effectively by producing triangular shapes internally. This can produce several orders of magnitude of improvement on convex one-dimensional regression problems. Much more theoretical work is needed to realize its full potential beyond this context, but the implementation in this repository will still work in arbitrary numbers of dimensions. The file Triangle_Method.py is a generalized form of the method that will build each neuron its own custom 1-d convex activation function (with exponential efficiency). Example usage on one dimensional problems can be found in Example_Usage.ipynb and an example of using this in a real neural network can be found in Example_VGG16_CIFAR10.ipynb.

Milkert, Max [National Renewable Energy Laboratory↗

Vibration localization in mono- and bi-coupled bladed disks - A transfer matrix approach

A transfer matrix approach to the analysis of the dynamics of mistuned bladed disks is presented. The study focuses on mono-coupled systems, in which each blade is coupled to its two neighboring blades, and bi-coupled systems, where each blade is coupled to its four nearest neighbors. Transfer matrices yield the free dynamics, both the characteristic free wave and the normal modes - in closed form for the tuned assemblies. Mistuned assemblies are represented by random transfer matrices and an examination of the effect of mistuning on harmonic wave propagation yields the localization factor - the average rate of spatial wave amplitude decay per blade - in the mono-coupled assembly. Based on a comparison of the wave propagation characteristics of the mono- and bi-coupled assemblies, important conclusions are drawn about the effect of the additional coupling coordinate on the sensitivity to mistuning and the strength of mode localization predicted by a mono-coupled analysis.

Ottarsson, Gisli↗

A microbenchmark characterization of the Emu chick

The Emu Chick is a prototype system designed around the concept of migratory memory-side processing. Rather than transferring large amounts of data across power-hungry, high-latency interconnects, the Emu Chick moves lightweight thread contexts to near-memory cores before the beginning of each memory read. The current prototype hardware uses FPGAs to implement cache-less “Gossamer” cores for computational work and rely on a typical stationary core (PowerPC) to run basic operating system functions and migrate threads between nodes. In this multi-node characterization of the Emu Chick, we extend an earlier single-node investigation of the the memory bandwidth characteristics of the system through benchmarks like STREAM, pointer chasing, and sparse matrix-vector multiplication. Here, we compare the Emu Chick hardware to architectural simulation and an Intel Xeon-based platform. Our results demonstrate that for many basic operations the Emu Chick can use available memory bandwidth more efficiently than a more traditional, cache-based architecture although bandwidth usage suffers for computationally intensive workloads like SpMV. Moreover, the Emu Chick provides stable, predictable performance with up to 65% of the peak bandwidth utilization on a random-access pointer chasing benchmark with weak locality.

97 MATHEMATICS AND COMPUTING↗

Magnetic Relaxation Experiments in CNT-Based Magnetic Nanocomposite

In this work, we discuss the relaxation of the magnetic moments in a novel carbon nanotube (CNT)-based nanocomposite synthesized by using chemical vapor deposition process. The material consists of a matrix of CNT filled by Fe-based nanoparticles. This structure is seen clearly by scanning and transmission. X-ray diffraction and Raman spectroscopy are used to detect the predominant Fe{sub 3}C phase and the CNT presence in the sample, respectively. The results obtained from both hysteresis cycles, M(H), and zero field cooled-field cooled (ZFC-FC) measurements confirm that the material is characterized by both a strong ferromagnetic exchange and random magnetic anisotropy. For the first time, we have been able to fit the magnetic relaxation data, M(t), by using both the two distributions of nanoparticles data deduced from the ZFC-FC data and the temperature dependence of the magnetic anisotropy obtained from the law of approach to saturation in random magnets. .

77 NANOSCIENCE AND NANOTECHNOLOGY↗

Progress on Complex Langevin simulations of a finite density matrix model for QCD

We study the Stephanov model, which is an RMT model for QCD at finite density, using the Complex Langevin algorithm. Naive implementation of the algorithm shows convergence towards the phase quenched or quenched theory rather than to intended theory with dynamical quarks. A detailed analysis of this issue and a potential resolution of the failure of this algorithm are discussed. We study the effect of gauge cooling on the Dirac eigenvalue distribution and time evolution of the norm for various cooling norms, which were specifically designed to remove the pathologies of the complex Langevin evolution. The cooling is further supplemented with a shifted representation for the random matrices. Unfortunately, none of these modifications generate a substantial improvement on the complex Langevin evolution and the final results still do not agree with the analytical predictions.

Bloch, Jacques↗

Synthesis, Characterization and Properties of Nanoparticles of Intermetallic Compounds

The research program from 2010 to the end of the grant focused on understanding the factors important to the synthesis of single phase intermetallic nano-particles (NPs), their size, crystalline order, surface properties and electrochemical activity. The synthetic method developed is a co-reduction of mixtures of single metal precursors by strong, soluble reducing agents in a non-protic solvent, tetrahydrofuran (THF). With some exceptions, the particles obtained by room temperature reduction are random alloys that need to be annealed at modest temperatures (200 to 600 °C) in order to develop an ordered structure. To avoid significant particle size growth and agglomeration, the particles must be protected by surface coatings. We developed a novel method of coating the metal nanoparticles with KCl, a by-product of the reduction reaction if the proper reducing agents are employed. In that case, a composite product containing individual metal nanoparticles in a KCl matrix is obtained. The composite can be heated to at least 600 °C without significant agglomeration or growth in particle size. Washing the annealed product in the presence of catalyst supports in ethylene glycol removes the KCl and deposits the particles on the support. Six publications present the method and its application to producing and studying new catalyst/support combinations for fuel cell applications. Three publications concern the use of related methods to explore new lithium-sulfur battery concepts.

77 NANOSCIENCE AND NANOTECHNOLOGY↗

Electrical-current-induced magnetic hysteresis in self-assembled vertically aligned L a 2 / 3 S r 1 / 3 Mn O 3 : ZnO nanopillar composites

Magnetoresistive random-access memory (MRAM) is poised to become a next-generation information storage device. Yet, many materials challenges remain unsolved before it can become a widely used memory storage solution. Among them, an urgent need is to identify a material system that is suitable for downscaling and is compatible with low-power logic applications. Self-assembled, vertically aligned La 2/3 Sr 1/3 MnO 3 : ZnO nanocomposites, in which La 2/3 Sr 1/3 MnO 3 (LSMO) matrix and ZnO nanopillars form an intertwined structure with coincident-site-matched growth occurring between the LSMO and ZnO vertical interfaces, may offer new MRAM applications by combining their superior electric, magnetic ( B ), and optical properties. Here, in this Rapid Communication, we show the results of electrical current induced magnetic hysteresis in magnetoresistance measurements in these nanopillar composites. We observe that when the current level is low, for example, 1 µA, the magnetoresistance displays a linear, negative, nonhysteretic B field dependence. Surprisingly, when a large current is used, I > 10 µA, a hysteretic behavior is observed when the B field is swept in the up and down directions. This hysteresis weakens as the sample temperature is increased. Finally, a possible spin-valve mechanism related to this electrical current induced magnetic hysteresis is proposed and discussed.

36 MATERIALS SCIENCE↗