Search NASA⌕ Search

SEARCH · Search NASA

Results for “distributed clustering 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 703 records · Page 39

THE SPINE OF THE COSMIC WEB

We present the SpineWeb framework for the topological analysis of the Cosmic Web and the identification of its walls, filaments, and cluster nodes. Based on the watershed segmentation of the cosmic density field, the SpineWeb method invokes the local adjacency properties of the boundaries between the watershed basins to trace the critical points in the density field and the separatrices defined by them. The separatrices are classified into walls and the spine, the network of filaments and nodes in the matter distribution. Testing the method with a heuristic Voronoi model yields outstanding results. Following the discussion of the test results, we apply the SpineWeb method to a set of cosmological N-body simulations. The latter illustrates the potential for studying the structure and dynamics of the Cosmic Web.

79 ASTRONOMY AND ASTROPHYSICS↗

A Parallel Solver for Graph Laplacians

Problems from graph drawing, spectral clustering, network flow and graph partitioning can all be expressed in terms of graph Laplacian matrices. There are a variety of practical approaches to solving these problems in serial. However, as problem sizes increase and single core speeds stagnate, parallelism is essential to solve such problems quickly. We present an unsmoothed aggregation multigrid method for solving graph Laplacians in a distributed memory setting. We introduce new parallel aggregation and low degree elimination algorithms targeted specifically at irregular degree graphs. These algorithms are expressed in terms of sparse matrix-vector products using generalized sum and product operations. This formulation is amenable to linear algebra using arbitrary distributions and allows us to operate on a 2D sparse matrix distribution, which is necessary for parallel scalability. Our solver outperforms the natural parallel extension of the current state of the art in an algorithmic comparison. We demonstrate scalability to 576 processes and graphs with up to 1.7 billion edges.

Konolige, Tristan↗

NMR_DIFFSIM v. 1.0

The software generates and analyzes nuclear magnetic resonance spin diffusion simulations for the purpose of measuring domain sizes or spin diffusion constants in a heterogenous sample. Its primary capabilities are: 1. It solves the spin magnetization diffusion equation to simulate the results of a spin diffusion experiment using any structural input geometry, including structure models generated by molecular dynamics or other computational or experimental methods. 2. It compares the results of simulations on an input structure to experimental data, providing a data fitting algorithm to extract physical information from the experimental data via the simulation. 3. It can provide statistical measures to compare correlations between structural models, to determine domain clustering and connectedness, and to provide goodness-of-fit measures including chord length distributions, chi-squared values, Pearson coefficients and pair correlation functions. 4. It provides output reports, data files, and movies to visualize simulation results.

Sorte, Eric↗

Constraints on the formation history of the elliptical galaxy NGC 3923 from the colors of its globular clusters

We present a study of the colors of globular clusters associated with the elliptical galaxy NGC 3923. Our final sample consists of Wasington system C and T(sub 1) photometry for 143 globular cluster candidates with an expected contamination of no more than 10%. We find that the color distribution of the NGC 3923 globular cluster system (GCS) is broad and appears to have at least two peaks. A mixture modeling analysis of the color distribution indicates that a two-component model is favored over a single-component one at a high level of confidence (greater than 99%). This evidence for more than one population in the GCS of NGC 3923 is similar to that previously noted for the four other elliptical galaxies for which similar data have been published. Furthermore, we find that the NGC 3923 GCS is redder than the GCSs of previously studed elliptical galaxies of similar luminosity. The median metallicity inferred from our (C-(T(sub 1)))(sub 0) colors is (Fe/H)(sub med) = -0.56, with an uncertainty of 0.14 dex arising from all sources of uncertainty in the mean color. This is more metal rich than the median metallicity found for the GCS of M87 using the same method, (Fe/H)(sub med) = -0.94. Since M87 is more luminous than NGC 3923, this result points to significant scatter about any trend of higher GCS metallicity with increasing galaxy luminosity. We also show that there is a color gradient in the NGC 3923 GCS corresponding to about -0.5 dex in Delta(Fe/H)/Delta(log r). We conclude that the shape of the color distribution of individual GCSs and the variation in mean color among the GCSs of ellipticals are difficult to understand if elliptical galaxies are formed in a single protogalactic collapse. Models in which ellipticals and their globular clusters are formed in more than one event, such as a merger scenario, are more successful in accounting for these observations.

Zepf, Stephen E.↗

On the state of Mn in Mn{sub x}Zn{sub 1−x}O nanoparticles and their surface modification with isonipecotic acid

Mn-doped ZnO (Mn{sub x}Zn{sub 1−x}O) nanoparticles were synthesized by the co-precipitation method and coated with isonipecotic acid as capping ligand. The structure, composition and morphology of the resulting nanomaterial were investigated by energy disperse X-ray analysis, X-ray diffraction, and transmission electron microscopy data. Such measurements showed that the solid obtained contains 6 at% of Mn and it is formed by a highly crystalline material with 3–5 nm range of crystallite size, and only a small elongation of its cell parameter with respect to undoped ZnO wurtzite unit cell. Information on the state of manganese atom in the Mn{sub x}Zn{sub 1−x}O nanostructures formed was obtained from X-ray photoelectron (XPS) and electron energy loss (EELS) spectroscopies. XPS and EELS spectra are composed of four peaks, corresponding to two species of Mn(II) and signals from Mn(III) and Mn(IV). Such spectral data on the state of Mn in the material studied is consistent with the mapping of Mn distribution observed in recorded transmission electron microscopy images, which reveal presence of clusters of Mn atoms. Only a fraction of doping Mn atoms were found forming a solid solution with the host ZnO structure. The functionalization of the nanoparticles system with Isonipecotic acid shows that this molecule remains anchored to the nanoparticles surface mainly through its N basic site. The availability of free carboxylate groups in the capping molecule was tested by conjugation to type IV horseradish peroxidase. - Graphical abstract: State of Mn atoms in Mn-doped ZnO nanostructures prepared by the precipitation method, their capping with isonipecotic acid and subsequent conjugation to peroxidase. - Highlights: • State of manganese in manganese-doped zinc oxide nanoparticles. • Isonipecotic acid as surface modifier of ZnO nanoparticles. • Peroxidase conjugation to ZnO nanoparticles modified with isonipecotic acid.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Water and chloride ions migration in porous cementitious materials: An experimental and molecular dynamics investigation

The interactions of water and chloride ions penetration into porous mortar have been investigated experimentally and modelled by molecular dynamics. The water distribution visualized by neutron radiography and chloride penetration by Electron Probe Micro-Analysis indicate that separation of chloride ions from water during the penetration process took place and water ran much deeper into the material than chloride ions. The relationship between the penetration depth of water and chloride ions can be described by a linear function despite different water-cement ratio. Furthermore, molecular dynamics attributes the different transport mechanisms of water and ions in porous cementitious materials to three reasons: the nanostructure of hydration product play filtering role to screen chloride ions with larger hydration shell; as compared with water, ions reside in the calcium silicate surface for longer time due to the strong interfacial bonding; chloride ions accumulate and form CaCl ionic cluster, slowing down the migrating rate.

36 MATERIALS SCIENCE↗

miniVite: A Graph Analytics Benchmarking Tool for Massively Parallel Systems

Benchmarking of high performance computing sys-tems can help provide critical insights for ef?cient design of computing systems and software applications. Although a large number of tools for benchmarking exist, there is a lack of rep-resentative benchmarks for the class of irregular computations as exempli?ed by graph analytics. In this paper, we propose miniVite as a representative graph analytics benchmark tool to test a variety of distributed-memory systems. Graph clustering, popularly known as community detection, is a prototypical graph operation used in numerous scienti?c computing and analytics applications. The goal of clustering is to partition a graph into clusters (or communities) such that each cluster consists of vertices that are densely connected within the cluster and sparsely connected to the rest of the graph. Modu-larity optimization is a popular technique for identifying clusters in a graph. Ef?cient parallelization of modularity optimization-based algorithms is challenging. One successful approach was conceived in Vite, a distributed-memory implementation of the Louvain algorithm that incorporates several heuristics. We introduce miniVite as a representative but simpli?ed variant of Vite, to serve as a prototypical graph analytics benchmarking tool. Unlike other graph-based methods such as breadth-?rst search and betweenness centrality, miniVite represents highly complex computational patterns stressing a variety of system features, which can provide crucial insight for co-design of future computing systems.

community detection, parallel, MPI↗

SU-E-J-03: A Comprehensive Comparison Between Alpha and Beta Emitters for Cancer Radioimmunotherapy

Purpose: The purpose of this study is to perform a comprehensive comparison of the therapeutic efficacy and cytotoxicity of alpha and beta emitters for Radioimmunotherapy (RIT). For each stage of cancer development, specific models were built for the separate objectives of RIT to be addressed:a) kill isolated cancer cells in transit in the lymphatic and vascular circulation,b) regress avascular cell clusters,c) regress tumor vasculature and tumors. Methods: Because of the nature of short range, high LET alpha and long energy beta radiation and heterogeneous antigen expression among cancer cells, the microdosimetric approach is essential for the RIT assessment. Geant4 based microdosimetric models are developed for the three different stages of cancer progression: cancer cells, cell clusters and tumors. The energy deposition, specific energy resulted from different source distribution in the three models was calculated separately for 4 alpha emitting radioisotopes ({sup 211}At, {sup 213}Bi, {sup 223}Ra and {sup 225}Ac) and 6 beta emitters ({sup 32}P, {sup 33}P, {sup 67}Cu, {sup 90}Y, {sup 131}I and {sup 177}Lu). The cell survival, therapeutic efficacy and cytotoxicity are determined and compared between alpha and beta emitters. Results: We show that internal targeted alpha radiation has advantages over beta radiation for killing isolated cancer cells, regressing small cell clusters and also solid tumors. Alpha particles have much higher dose specificity and potency than beta particles. They can deposit 3 logs more dose than beta emitters to single cells and solid tumor. Tumor control probability relies on deep penetration of radioisotopes to cancer cell clusters and solid tumors. Conclusion: The results of this study provide a quantitative understanding of the efficacy and cytotoxicity of RIT for each stage of cancer development.

60 APPLIED LIFE SCIENCES↗

MID-INFRARED EXTINCTION MAPPING OF INFRARED DARK CLOUDS: PROBING THE INITIAL CONDITIONS FOR MASSIVE STARS AND STAR CLUSTERS

Infrared dark clouds (IRDCs) are cold, dense regions of giant molecular clouds that are opaque at wavelengths {approx}10 {mu}m or more and thus appear dark against the diffuse Galactic background emission. They are thought to be the progenitors of massive stars and star clusters. We use 8 {mu}m imaging data from Spitzer Galactic Legacy Mid-Plane Survey Extraordinaire to make extinction maps of 10 IRDCs, selected to be relatively nearby and massive. The extinction mapping technique requires construction of a model of the Galactic IR background intensity behind the cloud, which is achieved by correcting for foreground emission and then interpolating from the surrounding regions. The correction for foreground emission can be quite large, up to {approx}50% for clouds at {approx}5 kpc distance, thus restricting the utility of this technique to relatively nearby clouds. We investigate three methods for the interpolation, finding systematic differences at about the 10% level, which, for fiducial dust models, corresponds to a mass surface density {sigma} = 0.013 g cm{sup -2}, above which we conclude that this extinction mapping technique attains validity. We examine the probability distribution function of {sigma} in IRDCs. From a qualitative comparison with numerical simulations of astrophysical turbulence, many clouds appear to have relatively narrow distributions suggesting relatively low (less than five) Mach numbers and/or dynamically strong magnetic fields. Given cloud kinematic distances, we derive cloud masses. Rathborne, Jackson, and Simon identified cores within the clouds and measured their masses via millimeter dust emission. For 43 cores, we compare these mass estimates with those derived from our extinction mapping, finding good agreement: typically factors of {approx}<2 difference for individual cores and an average systematic offset of {approx}<10% for the adopted fiducial assumptions of each method. We find tentative evidence for a systematic variation of these mass ratios as a function of core density, which is consistent with models of ice mantle formation on dust grains and subsequent grain growth by coagulation, and/or with a temperature decrease in the densest cores.

99 GENERAL AND MISCELLANEOUS↗

Mutual diffusion of binary liquid mixtures containing methanol, ethanol, acetone, benzene, cyclohexane, toluene, and carbon tetrachloride

Mutual diffusion coefficients of all 20 binary liquid mixtures that can be formed out of methanol, ethanol, acetone, benzene, cyclohexane, toluene, and carbon tetrachloride without a miscibility gap are studied at ambient conditions of temperature and pressure in the entire composition range. The considered mixtures show a varying mixing behavior from almost ideal to strongly non-ideal. Predictive molecular dynamics simulations employing the Green-Kubo formalism are carried out. Radial distribution functions are analyzed to gain an understanding of the liquid structure influencing the diffusion processes. It is shown that cluster formation in mixtures containing one alcoholic component has a significant impact on the diffusion process. The estimation of the thermodynamic factor from experimental vapor-liquid equilibrium data is investigated, considering three excess Gibbs energy models, i.e., Wilson, NRTL, and UNIQUAC. It is found that the Wilson model yields the thermodynamic factor that best suits the simulation results for the prediction of the Fick diffusion coefficient. Four semi-empirical methods for the prediction of the self-diffusion coefficients and nine predictive equations for the Fick diffusion coefficient are assessed and it is found that methods based on local composition models are more reliable. Finally, the shear viscosity and thermal conductivity are predicted and in most cases favorably compared with experimental literature values.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

BigNeuron: a resource to benchmark and predict performance of algorithms for automated tracing of neurons in light microscopy datasets

BigNeuron is an open community bench-testing platform with the goal of setting open standards for accurate and fast automatic neuron tracing. We gathered a diverse set of image volumes across several species that is representative of the data obtained in many neuroscience laboratories interested in neuron tracing. Here, we report generated gold standard manual annotations for a subset of the available imaging datasets and quantified tracing quality for 35 automatic tracing algorithms. The goal of generating such a hand-curated diverse dataset is to advance the development of tracing algorithms and enable generalizable benchmarking. Together with image quality features, we pooled the data in an interactive web application that enables users and developers to perform principal component analysis, t-distributed stochastic neighbor embedding, correlation and clustering, visualization of imaging and tracing data, and benchmarking of automatic tracing algorithms in user-defined data subsets. The image quality metrics explain most of the variance in the data, followed by neuromorphological features related to neuron size. Furthermore, we observed that diverse algorithms can provide complementary information to obtain accurate results and developed a method to iteratively combine methods and generate consensus reconstructions. The consensus trees obtained provide estimates of the neuron structure ground truth that typically outperform single algorithms in noisy datasets. However, specific algorithms may outperform the consensus tree strategy in specific imaging conditions. Finally, to aid users in predicting the most accurate automatic tracing results without manual annotations for comparison, we used support vector machine regression to predict reconstruction quality given an image volume and a set of automatic tracings.

97 MATHEMATICS AND COMPUTING↗

The clustering of DESI-like luminous red galaxies using photometric redshifts

ABSTRACT We present measurements of the redshift-dependent clustering of a DESI-like luminous red galaxy (LRG) sample selected from the Legacy Survey imaging data set, and use the halo occupation distribution (HOD) framework to fit the clustering signal. The photometric LRG sample in this study contains 2.7 million objects over the redshift range of 0.4 < z < 0.9 over 5655 deg2. We have developed new photometric redshift (photo-z) estimates using the Legacy Survey DECam and WISE photometry, with σNMAD = 0.02 precision for LRGs. We compute the projected correlation function using new methods that maximize signal-to-noise ratio while incorporating redshift uncertainties. We present a novel algorithm for dividing irregular survey geometries into equal-area patches for jackknife resampling. For a five-parameter HOD model fit using the MultiDark halo catalogue, we find that there is little evolution in HOD parameters except at the highest redshifts. The inferred large-scale structure bias is largely consistent with constant clustering amplitude over time. In an appendix, we explore limitations of Markov chain Monte Carlo fitting using stochastic likelihood estimates resulting from applying HOD methods to N-body catalogues, and present a new technique for finding best-fitting parameters in this situation. Accompanying this paper, we have released the Photometric Redshifts for the Legacy Surveys catalogue of photo-z’s obtained by applying the methods used in this work to the full Legacy Survey Data Release 8 data set. This catalogue provides accurate photometric redshifts for objects with z < 21 over more than 16 000 deg2 of sky.

79 ASTRONOMY AND ASTROPHYSICS↗

The Binary INformation from Open Clusters Using SEDs (BINOCS) Project: Reliable Photometric Mass Determinations of Binary Star Systems in Clusters

We introduce a new binary detection technique, Binary INformation from Open Clusters using SEDs (binocs), which we show is able to determine reliable stellar multiplicity and masses over a much larger mass range than current approaches. This new technique determines accurate component masses of binary and single systems of the open clusters’ main sequence by comparing observed magnitudes from multiple photometric filters to synthetic star spectral energy distributions (SEDs), allowing us to systematically probe the binary population for low-mass stars in clusters for eight well-studied open clusters. We provide new deep, infrared photometric catalogs (1.2–8.0 μm) for the key open clusters NGC 1960 (M36), NGC 2099 (M37), NGC 2420, and NGC 2682 (M67), using observations from NOAO/NEWFIRM and Spitzer/IRAC. Using these deep multiwavelength catalogs, the binocs method is applied to these clusters to determine accurate component masses for unresolved cluster binaries. We explore binary fractions as a function of cluster age, Galactic location, and metallicity.

47 OTHER INSTRUMENTATION↗

The evolution of large-scale structure in a universe dominated by cold dark matter

The results of numerical simulations of nonlinear gravitational clustering in universes dominated by weakly interacting, 'cold' dark matter are presented. The numerical methods used and the way in which initial conditions were generated are described, and the simulations performed are catalogued. The evolution of the fundamental statistical properties of the models is described and their comparability with observation is discussed. Graphical comparisons of these open models with the observed galaxy distribution in a large redshift survey are made. It is concluded that a model with a cosmological density parameter omega equal to one is quite unacceptable if galaxies trace the mass distribution, and that models with omega of roughly two, while better, still do not provide a fully acceptable match with observation. Finally, a situation in which galaxy formation is suppressed except in sufficiently dense regions is modelled which leads to models which can agree with observation quite well even for omega equal to one.

Davis, M.↗

NASA Tech Briefs, July 2013

Dielectrophoresis-Based Particle Sensor Using Nanoelectrode Arrays; Multi-Dimensional Damage Detection for Surfaces and Structures; ULTRA: Underwater Localization for Transit and Reconnaissance Autonomy; Autonomous Cryogenic Leak Detector for Improving Launch Site Operations; Submillimeter Planetary Atmospheric Chemistry Exploration Sounder; Method for Reduction of Silver Biocide Plating on Metal Surfaces; Silicon Micromachined Microlens Array for THz Antennas; Forward-Looking IED Detector Ground Penetrating Radar; Fully Printed, Flexible, Phased Array Antenna for Lunar Surface Communication, Battery Charge Equalizer with Transformer Array; An Efficient, Highly Flexible Multi-Channel Digital Downconverter Architecture; Dimmable Electronic Ballast for a Gas Discharge Lamp; Conductive Carbon Nanotube Inks for Use with Desktop Inkjet Printing Technology; Enhanced Schapery Theory Software Development for Modeling Failure of Fiber-Reinforced Laminates; High-Performance, Low-Temperature-Operating, Long-Lifetime Aerospace Lubricants; Carbon Nanotube Microarrays Grown on Nanoflake Substrates; Differential Muon Tomography to Continuously Monitor Changes in the Composition of Subsurface Fluids; Microgravity Drill and Anchor System; 20 Granular Media-Based Tunable Passive Vibration Suppressor; 21 Miga Aero Actuator and 2D Machined Mechanical Binary Latch; Micro-XRF for In Situ Geological Exploration of Other Planets; Hydrogen-Enhanced Lunar Oxygen Extraction and Storage Using Only Solar Power; Uplift of Ionospheric Oxygen Ions During Extreme Magnetic Storms; Miniaturized, High-Speed, Modulated X-Ray Source; Hollow-Fiber Spacesuit Water Membrane Evaporator 25 High-Power Single-Mode 2.65-micrometers InGaAsSb/AlInGaAsSb Diode Lasers; Optical Device for Converting a Laser Beam Into Two Co-aligned but Oppositely Directed Beams; A Hybrid Fiber/Solid-State Regenerative Amplifier with Tunable Pulse Widths for Satellite Laser Ranging; X-Ray Diffractive Optics; SynGenics Optimization System (SynOptSys); 29 CFD Script for Rapid TPS Damage Assessment; radEq Add-On Module for CFD Solver Loci-CHEM; Science Opportunity Analyzer (SOA) Version 8; 30 Autonomous Byte Stream Randomizer; Distributed Engine Control Empirical/Analytical Verification Tools; Dynamic Server-Based KML Code Generator Method for Level-of-Detail Traversal of Geospatial Data; Automated Planning of Science Products Based on Nadir Overflights and Alerts for Onboard and Ground Processing; Linked Autonomous Interplanetary Satellite Orbit Navigation; Risk-Constrained Dynamic Programming for Optimal Mars Entry, Descent, and Landing; Scheduling Operations for Massive Heterogeneous Clusters; Deepak Condenser Model (DeCoM); Flight Software Math Library; Recirculating 1-K-Pot for Pulse-Tube Cryostats; 35 Method for Processing Lunar Regolith Using Microwaves; Wells for In Situ Extraction of Volatiles from Regolith (WIEVR); and Estimating the Backup Reaction Wheel Orientation Using Reaction Wheel Spin Rates Flight Telemetry from a Spacecraft.

Source record↗

Retrieve Methane from IR sounder measurements Using Machine Learning-Enhanced Physical Inversion

The sensitivity of IR sounder measurements to atmospheric CH 4 is often limited due to interferences from signals of other trace gases, insufficient thermal contrast, and cloud blockage. In order to resolve the geographical and vertical distribution of atmospheric CH 4 profiles, accurate scene-dependent a priori information is critically needed to support an optimal estimation method-based physical inversion scheme. Following the principles of indexing, representation, and retrieval, a spectral fingerprinting methodology is developed to address the needs for both accuracy and computational efficiency in sounder-based CH 4 retrieval. Within this framework, a clustering method based on machine learning is first employed to stratify and identify the a priori state within the pre-constructed database, using optimized spectral radiances as predictors. The corresponding radiative kernel is then used to establish the physical inversion scheme for finding the solution. High-quality data from CH 4 data assimilation systems like the Carbon-Tracker and the Copernicus Atmosphere Monitoring Service (CAMS) reanalysis, as well as the state-of-art sounder products are used to build the training database, including radiative kernels. We will demonstrate the results retrieved from CrIS observations and the associated validation work.

Wan Wu↗

Dynamic structural disorder in supported nanoscale catalysts

We investigate the origin and physical effects of “dynamic structural disorder” (DSD) in supported nano-scale catalysts. DSD refers to the intrinsic fluctuating, inhomogeneous structure of such nano-scale systems. In contrast to bulk materials, nano-scale systems exhibit substantial fluctuations in structure, charge, temperature, and other quantities, as well as large surface effects. The DSD is driven largely by the stochastic librational motion of the center of mass and fluxional bonding at the nanoparticle surface due to thermal coupling with the substrate. Our approach for calculating and understanding DSD is based on a combination of real-time density functional theory/molecular dynamics simulations, transient coupled-oscillator models, and statistical mechanics. This approach treats thermal and dynamic effects over multiple time-scales, and includes bond-stretching and -bending vibrations, and transient tethering to the substrate at longer ps time-scales. Potential effects on the catalytic properties of these clusters are briefly explored. Model calculations of molecule-cluster interactions and molecular dissociation reaction paths are presented in which the reactant molecules are adsorbed on the surface of dynamically sampled clusters. This model suggests that DSD can affect both the prefactors and distribution of energy barriers in reaction rates, and thus can significantly affect catalytic activity at the nano-scale.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Energetic recoils in UO2 simulated using five different potentials

This report presents the results of classical molecular dynamics simulations of the diffuse pre-melting transition, melting, and defect production by 1 keV U recoils in UO2 using five different rigid-ion potentials. The experimentally-observed pre-melting transition occurred for all five cases. For all the potentials studied, dynamic defect annealing is highly effective and is accompanied by replacement events on the anion sublattice. The primary damage state after ~15 ps consists of isolated Frenkel pairs and interstitial and vacancy clusters of various sizes. The average displacement energy varies from ~28 to ~83 eV and the number of Frenkel pairs is different by a factor of three depending on the choice of potential. The size and spatial distribution of vacancy and interstitial clusters is drastically different for the potentials studied. The results provide statistics of defect production. They point to a pressing need to determine defect formation, migration and binding energies in UO2 from first principles and to develop reliable potentials based on this data for simulating microstructural evolution in nuclear fuel under operating conditions.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗