SEARCH · Search NASA
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.
Raman scattering as a technique for studying the formation of tungsten silicide is presented. The tungsten silicide films have been formed by rapid thermal annealing of thin tungsten films sputter deposited on silicon substrates. The Raman data are interpreted by using data from resistivity measurements, Auger and Rutherford backscattering measurements, and scanning electron microscopy.
Carbon fibers have been grown from methane gas, on iron seeded substrates at 1370 K, and subsequently annealed to a series of temperatures between 2800 K and 3475 K. The subsequent measurement of the electrical resistivity are reported as a function of temperature from 4 K to 2100 K. The two band model for the resistivity of graphite is well fit to the data, and results in physically reasonable parameters of the fits. The resistivity vs temperature results for two boron doped fibers are reported as well. The product of the resistivity times density for these fibers is lower than that product for any refractory metal, for temperatures above about 1000 K, suggesting the usefulness of these fibers as high temperature materials in space applications.
Self-organization and learning is a distinctive feature of neural nets and processors that sets them apart from conventional approaches to signal processing. It leads to self-programmability which alleviates the problem of programming complexity in artificial neural nets. In this paper architectures for partitioning an optoelectronic analog of a neural net into distinct layers with prescribed interconnectivity pattern to enable stochastic learning by simulated annealing in the context of a Boltzmann machine are presented. Stochastic learning is of interest because of its relevance to the role of noise in biological neural nets. Practical considerations and methodologies for appreciably accelerating stochastic learning in such a multilayered net are described. These include the use of parallel optical computing of the global energy of the net, the use of fast nonvolatile programmable spatial light modulators to realize fast plasticity, optical generation of random number arrays, and an adaptive noisy thresholding scheme that also makes stochastic learning more biologically plausible. The findings reported predict optoelectronic chips that can be used in the realization of optical learning machines.
Thermoluminescence (TL) properties were measured in samples of four type-3.0 chondrites annealed at 400-850 C and 0.77-1 kbar in the presence of various amounts of water and sodium disilicate. Several changes recorded in TL characteristics, such as the lowering of TL sensitivity in certain samples, its increase in other samples, and changes in the peak position and peak width of TL suggested the occurrence of metamorphic processes in these samples. It is suggested that, for the chondrite types above 3.2, the observed changes in TL are consistent with the TL-metamorphism trends, being due to the formation of feldspar by the devitrification of chondrule glass during metamorphism. For types below 3.2, the TL data are consistent with the hypothesis that these chondrules experienced lower levels of metamorphism than the higher types or, alternatively, with the hypothesis that the type 3.0 chondrules are being produced from higher types by aqueous alteration.
Attention is given to network models in which each of the cones of the retina is given a provisional color at random, and then the cones are allowed to determine the colors of their neighbors through an iterative process. A symmetric-structure spin-glass model has allowed arrays to be generated from completely random arrangements of red and green to arrays with approximately as much disorder as the parafoveal cones. Simulated annealing has also been added to the process in an attempt to generate color arrangements with greater regularity and hence more revealing moirepatterns than than the arrangements yielded by quenched spin-glass processes. Attention is given to the perceptual implications of these results.
The optimal placement of discrete actuators and sensors is posed as a combinatorial optimization problem. Two examples for truss structures were used for illustration; the first dealt with the optimal placement of passive dampers along existing truss members, and the second dealt with the optimal placement of a combination of a set of actuators and a set of sensors. Except for the simplest problems, an exact solution by enumeration involves a very large number of function evaluations, and is therefore computationally intractable. By contrast, the simulated annealing heuristic involves far fewer evaluations and is best suited for the class of problems considered. As an optimization tool, the effectiveness of the algorithm is enhanced by introducing a number of rules that incorporate knowledge about the physical behavior of the problem. Some of the suggested rules are necessarily problem dependent.
The simulated annealing technique has been applied successfully to the problem of estimating the coefficients of a function in cases where only a portion of the data being fitted to the function is truly representative of the function, the rest being erroneous. Two examples are given, one in photometric function fitting and the other in pattern recognition. A schematic of the algorithm is provided.
The thermoelectric conversion efficiency of a GaP doped SiGe alloy is improved about 30 percent by annealing the alloy at a temperature above the melting point of the alloy, preferably stepwise from 1200 C to 1275 C in air to form large grains having a size over 50 microns and to form a GeGaP rich phase and a silicon rich phase containing SiP and SiO2 particles.
Y-Ba-Cu-O bulk materials prepared using the melt quench and post annealing method were shown to have very high maximum as well as remanent magnetization. Studies were carried out on materials prepared using this method which deviate from the Y:Ba:Cu = 1:2:3 stoichiometry. In one series of materials, only the stoichiometry was changed, in particular by introducing an excess of yttrium. In other cases, dopants including several rare earths were introduced. Effects of variations in composition on microstructure and phase evolution are discussed, as well as effects on the magnetic susceptibility and on the magnetization. The results show that doped materials can exhibit improvements in magnetic properties. Furthermore, the use of dopants sheds light on the role of defect sites in flux pinning.
By comparing IRAS mean colors and LRS spectral differences in a large sample of Mira variables with the chronological sequence for development of microwave maser emission among such stars, it appears possible to describe a dust grain formation and annealing sequence. The early time spectra are dominated by aluminum oxides, which are then overrun by the emergence of silicates. Rather than the growth of silicate mantles on aluminate cores, this sequence is interpreted as a simple result of the higher electron affinity of aluminum for oxygen, resulting in an initial abundance of AlO bonds in the underoxidized grains. Once all the Al becomes fully oxidized, the relative abundances of Al and Si dictate that the AlO signature will be swamped by the growth of the SiO stretching fundamental. The correlations among the proposed dust chronology, the increasingly thick circumstellar envelopes, and the light curve asymmetries of Mira variables are instructive for understanding the evolutionary changes occurring in such stars.
Inaccuracies in the length of members and the diameters of joints of large space structures may produce unacceptable levels of surface distortion and internal forces. Here, two discrete optimization problems are formulated, one to minimize surface distortion (DSQRMS) and the other to minimize internal forces (FSQRMS). Both of these problems are based on the influence matrices generated by a small-deformation linear analysis. Good solutions are obtained for DSQRMS and FSQRMS through the use of a simulated annealing heuristic.
Annealing procedures that greatly reduce linear birefringence in single-mode fiber coils are described. These procedures have been successfully applied to coils ranging from 5 mm to 10 cm in diameter and up to 200 or more turns. They involve temperature cycles that last 3-4 days and reach maximum temperatures of about 850 C. The residual birefringence and induced loss are minimized by proper selection of fiber. The primary application of these coils is optical fiber current sensors, where they yield small sensors that are more stable than those achieved by other techniques. A current sensor with a temperature stability of 8.4 x 10 to the -5th/K over the range from -75 to 145 C has been demonstrated. This is approximately 20 percent greater than the temperature dependence of the Verdet constant. Packaging degrades the stability, but a packaged sensor coil with a temperature stability of about 1.6 + 10 to the -4th/K over the range from -20 to 120 C has also been demonstrated.
The geometrical structures of M(+) (Ar)n ions, with n = 1-14, have been studied by the minimization of a many-body potential surface with a simulated annealing procedure. The minimization method is justified for finite systems through the use of an information theory approach. It is carried out for eight potential-energy surfaces constructed with two- and three-body terms parametrized from experimental data and ab initio results. The potentials should be representative of clusters of argon atoms with first-row transition-metal monocations of varying size. The calculated geometries for M(+) = Co(+) and V(+) possess radial shells with small (ca. 4-8) first-shell coordination number. The inclusion of an ion-induced-dipole-ion-induced-dipole interaction between argon atoms raises the energy and generally lowers the symmetry of the cluster by promoting incomplete shell closure. Rotational constants as well as electric dipole and quadrupole moments are quoted for the Co(+) (Ar)n and V(+) (Ar)n predicted structures.
The Shuttle Engineering Simulator (SES) is a facility which houses the software and hardware for a variety of simulation systems. The simulators include the Autonomous Remote Manipulator, the Manned Maneuvering Unit, Orbiter/Space Station docking, and shuttle entry and landing. The SES simulators are used by various groups throughout NASA. For example, astronauts use the SES to practice maneuvers with the shuttle equipment; programmers use the SES to test flight software; and engineers use the SES for design and analysis studies. Due to its high demand, the SES is busy twenty-four hours a day and seven days a week. Scheduling the facility is a problem that is constantly growing and changing with the addition of new equipment. Currently a number of small independent programs have been developed to help solve the problem, but the long-term answer lies in finding a flexible, integrated system that provides the user with the ability to create, optimize, and edit the schedule. COMPASS is an interactive and highly flexible scheduling system. However, until recently COMPASS did not provide any optimization features. This paper describes the simulated annealing extension to COMPASS. It now allows the user to interweave schedule creation, revision, and optimization. This practical approach was necessary in order to satisfy the operational requirements of the SES.
The effect of annealing the SCS-6 SiC fiber for one hour at 2000 C in an argon atmosphere is reported. The SiC grains in the fiber coarsen appreciably and the intergranular carbon films segregate to the grain junctions. It would appear that grain growth in the outer part of the fiber is primarily responsible for the loss in fiber strength and improvement in fiber creep resistance.
This paper presents preliminary results of an investigation on mistuning of bladed-disk assemblies aimed at capturing the benefits of mistuning on stability, while at the same time, minimizing the adverse effects on response by solving the following problem: given a set of N turbine blades, each being a small random perturbation of the same nominal blade, determine the best arrangement of the N blades in a mistuned cascade with regard to aeroelastic response. In the studies reported here, mistuning of the blades is restricted to small differences in torsional stiffness. The large combinatorial optimization problem of seeking the best arrangement by blade exchanges is solved using a simulated annealing algorithm.
For many decades the dramatic effect of trace amounts of reactive elements on alumina and chromia scale adhesion has been recognized and widely studied. Although various theories have been used to account for such behavior, the connection between scale adhesion and sulfur segregation was initially reported by Smeggil et al. This study found strong surface segregation of sulfur from very low levels in the bulk which could then be curtailed by the addition of reactive elements. It was assumed that the reactive elements, which are strong sulfide formers, acted by getting sulfur in the bulk thus precluding sulfur segregation and weakening of the oxide-metal bond. Subsequent studies confirmed that adhesion could be produced by reducing the sulfur impurity level, without reactive elements. The understanding of this phenomenon has been applied to modern single crystal superalloys, where the addition of Y, although very effective, is problematic. Also problematic is definition of the level of sulfur that is acceptable and below which no further adhesion benefit is reached. Published works have indicated a broad transition defined by various materials and oxidation tests. The present study describes the oxidation behavior of one superalloy (PWA 1480) as a function of various sulfur contents produced by hydrogen annealing for various temperatures, times, and sample thicknesses. The purpose is to define more precisely a criterion for adhesion based on total sulfur reservoir and segregation potential.
100 kHz magnetization properties of sample transverse magnetically annealed, cobalt-based amorphous and iron-based nanocrystalline tape wound magnetic cores are presented over the temperature range of -150 C to 150 C, at selected values of B(sub peak). Frequency resolved characteristics are given over the range of 50 kHz to 1 MHz, but at B(sub peak) = 0.1 T and 50 C only. Basic exciting winding current and induced voltage data were taken on bare toroidal cores, in a standard type measurement setup. A linear permeability model, which represents the core by a parallel L-R circuit, is used to interpret and present the magnetization characteristics and several figures of merit applicable to inductor materials are reviewed. The 100 kHz permeability thus derived decreases with increasing temperature for the Fe-based, nanocrystalline material, but increases roughly linearly with temperature for the two Co-based materials, as long as B(sub peak) is sufficiently low to avoid saturation effects. Due to the high permeabilities, rather low values of the 'quality factor' Q, from about 20 to below unity, were obtained over the frequency range of 50 kHz to 1 MHz (50 C, B(sub peak) = 0.1 T). Therefore these cores must be gapped in order to make up high Q or high current inductors. However, being rugged, low core loss materials with flat B-H loop characteristics, they may provide new solutions to specialty inductor applications.