Publications

Circumvention applications require a memory that retains data through radiation (total dose and transient) and loss of power. Various memory technologies have been reviewed and none, as yet, can meet these requirements. However, if complementary metal oxide silicon (CMOS) and silicon nitride oxide silicon (SNOS) memories are combined in a shadow RAM (random access memory) configuration, the requirements can be fulfilled. © 1991 IEEE

The use of carbon materials in catalytic system has been traditionally associated with their properties as supports. There are, however, some literature data describing their use as catalysts. Our results show that no simple relationship exists between the catalytic activity of some carbon materials for selective cleavage of carbon-carbon bonds adjacent to condensed polyaromatic rings an the elemental composition or surface area of these materials. The activity is dependent, however, upon the precursor used to generate the carbon material. Moreover, for a given precursor the activity is determined by the method of conversion to the carbon material. It remains to be established what elements of structure are responsible for the observed activity in carbon materials that are catalytically active. 12 refs., 7 tabs., 1 fig.

The strain rate in steady shock waves is proportional to the fourth power of shock amplitude for a wide variety of materials over a broad range of strain rates. A model based on this observation gives good agreement not only with steady-wave profiles but also with data on non-steady waves in aluminum. In apparent contrast, data on vanadium and uranium at low strain rates indicates a departure from the fourth power law if the wave profiles are assumed to be steady. However, when predicted profiles are produced by allowing the waves to propagate and evolve over the actual experimental sample thickness, the fourth power model gives excellent agreement with the wave profile data even though the wave profiles in the calculations have not yet reached steady state. The implication is that the experimental data do not represent steady waves, and the model is predicting the correct evolution of non-steady waves in vanadium and uranium. 7 refs., 2 figs.

The National Competitiveness Technology Transfer Act of 1989 has opened up the vast resources of our nation's national laboratories to the electronics industry. The electronics industry stands to gain advanced technology development, increased competitiveness, resource-sharing, and technology protection from this act. Sandia National Laboratories can help our nation's companies and universities in developing and applying advanced, commercially valuable technologies and in solving technological problems. These technological areas are discussed. A clear, non-bureaucratic process of tapping the microelectronics expertise and resources of Sandia National Laboratories is presented.

Highly reactive coal pyrites and unstable museum specimens are easily distinguished from the stable pyrites by the growth of white crystals that cover samples exposed to room atmosphere for short periods of time. Continued exposure to the atmosphere will eventually cause the specimens to fall apart. The term rotten pyrite has been applied to museum specimens that fall apart in this way. SEM studies show that reactive (rotten) pyrites contain between 100 and 10,000 times more dislocations than stable pyrites. Shock-loading of a stable pyrite to 7.5 GPa and 17 GPa increased its reactivity by a factor of two, probably caused by an increase in the number of imperfections. However, shock-loading at 22 GPa decreased the reactivity of pyrite because the imperfections produced at the higher pressure were removed during annealing the sample received at the higher temperature. Although there was a factor of six difference between the most and least reactive shocked MCB (commercial pyrite) samples, shock-loading did not increase the reactivity of the MCB pyrite to that of the Queensland coal pyrite. The results in hand show that while shock-loading produces sufficient imperfections to increase the reactivity of pyrites, there is insufficient data to show that imperfections are the main reason why some coal pyrites are highly reactive. 9 refs., 1 fig., 1 tab.

To implement the Qualified Manufacturers List (QML) approach to hardness assurance in a practical and cost-effective manner, one must identify technology parameters that affect radiation hardness and bring them under statistical process control. To aid this effort, we have developed a wafer-level test system to map test-structure and IC response across a wafer. This system permits current-voltage and charge-pumping measurements on transistors, and high-frequency capacitance-voltage measurements on capacitors. For frequencies up to 50 MHz, the system provides a complete menu of functional and parametric IC tests. Wafer maps and histograms of test-structure and IC response are presented for a 1.2-µm radiation-hardened CMOS technology to illustrate the capabilities of the wafer-level test system. Statistical and “deterministic” approaches to correlate test structure and IC response are discussed for this technology. © 1991 IEEE

Transmitting digital voice via packetized mobile communications systems that employ relatively short packet lengths and narrow bandwidths often necessitates very low bit rate coding of the voice data. Sandia National Laboratories is currently developing an efficient voice coding system operating at 800 bits per second (bps). The coding scheme is a modified version of the 2400 bps NSA LPC-10e standard. The most significant modification to the LPC-10e scheme is the vector quantization of the line spectrum frequencies associated with the synthesis filters. An outline of a hardware implementation for the 800 bps coder is presented. The speech quality of the coder is generally good, although speaker recognition is not possible. Further research is being conducted to reduce the memory requirements and complexity of the vector quantizer, and to increase the quality of the reconstructed speech. 4 refs., 2 figs., 3 tabs.

BA85 is a routine for the quantitative reduction of x-ray data collected from oxide samples in an electron microprobe. BA85 is based on the correction procedures developed by Bence and Albee and is coded in Flextran for use with the TASK8 microprobe operating system. Features include stoichiometry and statistical calculations, the use of a 90 - oxide A-factor matrix which contains all of the common valence states for such elements as Fe and Cr, the ability to analyze up to 45 oxides, and the ability to create and use mineral codes which permit associating up to 15 oxides with three letter mnemonic codes. Entering a mineral code results in the analysis of the oxides associated with it and the performance of one of 21 endmember calculations. 13 refs., 2 figs., 3 tabs.

This paper seeks to improve the synergism between computational aerodynamics and wind tunnel experimentation. In this paper, experimental and computational results are presented for a hypersonic vehicle configuration at Mach 8. Comparisons are made between experimental and computational results in order to improve the accuracy of both approaches. The basic vehicle configuration is a spherically blunted cone with a slice parallel with the axis of the vehicle. The half-angle of the cone is 10 deg. and the ratio of spherical nose radius to base radius in 10%. Onto the slice portion of the vehicle can be attached flaps with three different deflection angles; 10, 20, and 30 deg. All of the experimental results were obtained in the Sandia Mach 8 long duration, blow-down, hypersonic wind tunnel. Flow visualization results include surface oil flow, spark schlieren, and liquid crystal photographs and video. The liquid crystals were used as an aid in verifying that a laminar boundary layer existed over the entire body. An extensive uncertainty analysis was conducted to estimate quantitatively the accuracy of the measurement. Computational aerodynamic force and moment predictions are compared with the wind tunnel data. The Sandia Parabolized Navier-Stokes code is used to generate solutions for the sliced vehicle (no flap) and partial solutions for the flapped vehicle. For the geometry with the flap, an axially separated flow occurs and a time iterative Navier-Stokes code is used to provide comparisons with the data. This paper presents a portion of the results given in earlier works and also discusses new experimental results with this configuration.

Plotting and summary routines available for the TASK8 microprobe operating system are able to accept both spectral and quantitative data. All of the routines are able to be run as subroutines from within the TASK8 program or as stand alone programs. Additionally, the spectral plotting routine can be run from within a modified version of SQ. The quantitative routines currently in use with TASK8. Quantitative output can be sent by the summary program to a serial port that is connected to a VAX or PC in addition to printing it. The plotting codes have been written so that either a Tracor Northern TN2000 or a TN5xxx analyzer may be used with either a Hewlett Packard HP7221 series or a HP7470/HP7550 series plotter. The plotting routine for spectra incorporates a user definable usual was'' option to simplify most input procedures. The quantitative plotting routine offers numerous options including scale expansion, smoothing, auto-labeling, special symbols, and multiple pens. 5 refs., 5 figs.

A computational scheme is presented to solve the unsteady Navier-Stokes equations over a blunt body at high altitude, high Mach number atmospheric reentry flow conditions. This continuum approach is directed to low Reynolds/low density hypersonic flows by accounting for non-zero bulk viscosity effects in near frozen flow conditions. A significant difference from previous studies is the inclusion of the capability to model non-zero bulk viscosity effects. The grid definition for these low Reynolds number, viscous dominated flow fields is especially important in terms of numerical stability and accurate heat transfer solutions. 11 refs., 15 figs.

Difficulties in the accurate heat transfer computation of high speed, blunt body flows have been encountered by numerous researchers. The primary reason for these difficulties has been shown to be the grid dependency of the wall flux quantities. Obviously, the accuracy of the computed heat fluxes will, to a certain extent, depend on the particular numerical scheme employed. This article will be limited to the investigation of the flux vector splitting technique. An attempt has been made to develop procedures which will provide guidelines for selecting appropriate grid systems and, in particular, the grid line distribution near the surface for accurate heat transfer computations. The results have clearly shown the dependency of the heat flux quantities on the grid system. In addition, it is shown that changes in flow Mach number and/or Reynolds number may require further refinement of the grid system. 11 refs., 8 figs.

This paper describes a new technique for the detection of x-rays in electron column instruments used in microanalysis. In electron column instruments, the point source of x-rays is produced by the interaction of a focused electron beam with the sample. Neither of the conventional methods, wavelength dispersive (WDS) nor energy dispersive (EDS) based spectrometry, is optimized for low Z element quantitative analysis. In WDS applications, where the analyte elements are Be through P, chemical effects complicate the x-ray measurement process. Peak positions and shapes are altered, sometimes very strongly, by the electron configurations of the analyte atoms and neighboring atoms. In these cases, the ideal spectrometer would profile the peak and some small amount of continuum on either side of the peak such that an accurate peak area could be calculated. Present WDS spectrometers are serial in nature and cannot directly measure peak areas, often causing errors in the determination of light element concentrations. Bastin and co-workers have developed an elegant method to provide accurate area determinations, using the serial spectrometer, by a three point procedure. The parallel wavelength dispersive spectrometer (PWDS) we propose here is ideally suited for those applications.

A new facility is being constructed for the Primary Standards Laboratory at Sandia National Laboratories in Albuquerque, New Mexico. Salient features of the final design, described briefly in this paper, follow the conceptual design of 1984 to a fairly high degree. Because the facility is in process of construction, this paper is a progress report. 6 refs., 2 figs., 1 tab.

Curved and planar inversion domain boundaries (IDB) in aluminum nitride (AIN) form in sintered AIN ceramics containing oxygen, and oxygen is known to segregate to them. A number of interface models shown in Table 1, have been suggested based upon crystallographic constraints, chemical information and observed high resolution electron microscope (HREM) images. Until recently, problems with simulation of HREM images from AIN have made accurate determination of the structure of the IDB interface difficult. The aim of the present study was to use quantitative analytical electron microscopy (AEM) to determine the oxygen concentration at the IDBs, and then to compare the experimental results with calculated oxygen concentrations for each of the IDB models using a Monte Carlo electron trajectory simulation program. A match, if any, between the experimental and calculated oxygen concentrations would indicate the model which best described the IDB structure. The best match was obtained for Youngman's defect model. 14 refs., 5 figs., 3 tabs.

Plessite in iron meteorites is a two phase structure with an fcc precipitate phase in a bcc matrix. After Fe-Ni martensite forms during slow cooling, the martensite decomposition occurs at different temperatures. The morphology of the precipitates and the Ni content of both precipitate and matrix vary with the local average Ni composition of the plessite. In this study, the plessite structure of two octahedrites, Carlton and Grant, was characterized using the analytical electron microscope (AEM). The composition of the taenite precipitates in various regions of plessite which have 9 to 13 wt% and 15 to 20 wt% Ni composition were measured using an x-ray energy dispersive spectrometer (EDS) in the AEM. To understand the phase transformation processes which occurred during the plessite formation, an experimental set of Fe-Ni binary and Fe-Ni-P ternary alloys were made and analyzed also using the AEM. The alloys, which have 15 to 30 wt% Ni (0.2-0.3 wt% P for ternary alloys), were first homogenized at 1,200{degree}C and quenched to liquid nitrogen temperature to form martensite. They were then isothermally heat treated for 60 to 400 days in the temperature range from 450{degree}C to 300{degree}C. Two phase structures, which are similar to those of plessite, were formed in these alloys. The Fe-Ni phase equilibria measured in the decomposed martensite alloys can be used to explain the difference in Ni composition between precipitates in the high Ni and low Ni plessite regions. 3 refs., 2 figs.

Researchers around the world have demonstrated the effectiveness of titanium dioxide-based photocatalysis for decontaminating water containing hazardous organics and heavy metals. A great advantage of this process is that the organic is completely destroyed, leaving behind only water, carbon dioxide and dilute mineral acids. Also, the process has potential for doing two decontamination jobs at once: oxidizing organics while reducing toxic metals. As part of a program sponsored by the DOE, Sandia National Laboratories is carrying out large-scale tests to study the solar destruction of organics at realistic processing rates in addition to laboratory work aimed at determining the applicability of solar detoxification to the removal of heavy metals. In this paper, we present up-to-date results from Sandia's effort. The large-scale tests illustrate the effectiveness of solar detoxification for a variety of organics and compare measured reaction rates against published values of the rate constants for attack by aqueous hydroxyl radicals. This comparison highlights the importance of hydroxyl radical chemistry in solar detoxification processes. It is concluded that solar detoxification is only feasible for easily destroyed compounds like TCE, and PCE. The chlorinated methanes and ethanes are much more difficult to destroy and require very large solar collectors. In the second part of the paper, laboratory data are presented for a variety of different metals, including Ag, Cd, Cu, Hg, Ni, and Pt. The concentration of dissolved oxygen is one of the variables in the study. It is shown that the presence of dissolved metals can have a profound effect on organic oxidation rate, and that oxygen is not necessarily required for oxidation to occur. It is concluded that solar detoxification would be useful for removing mercury and silver but not for copper, nickel nor cadmium. 38 refs., 7 figs., 2 tabs.

Titanium dioxide (TiO{sub 2}) is a photocatalyst for solar detoxification of water containing organic contaminants such as solvents, PCB's, dioxins, pesticides, and dyes. Unfortunately, the ultraviolet (UV) energy used by TiO{sub 2} ({lambda}<400 nm) only comprises about 4% of the solar spectrum. One way of enhancing the efficiency of solar detoxification technologies is to utilize a larger portion of the solar spectrum to initiate the Tio{sub 2}- catalyzed detoxification chemistry. Metalloporphyrins strongly absorb visible and near infrared radiation. By utilization of a process called photosensitization, adsorption of these dyes onto TiO{sub 2} can enable a much broader portion of the solar spectrum to be used. Photosensitization relies upon the ability of the dye molecule to absorb more of the solar energy than bare TiO{sub 2} and to interact electronically with the TiO{sub 2} surface in such a way as to initiate TiO{sub 2}-based redox photochemistry using the dye-absorbed energy. 16 refs., 7 figs.

This document contains planned actions to correct the deficiencies identified in the Pre-Tiger Team Self-Assessment (PTTSA), January 1991, of Sandia National Laboratories (SNL -- Albuquerque, New Mexico; Tonopah, Nevada; and Kauai, Hawaii). The Self-Assessment was conducted by a Self-Assessment Working Group consisting of 19 department managers, with support from Environment, Safety, and Health (ES H) professionals, from October through December 1990. Findings from other past audits, dating back to 1985, were reviewed and compared with the PTTSA findings to determine if additional findings, key findings, or root causes were warranted. The resulting ES H Action Plan and individual planned actions were prepared by the ES H Action Plan Project Group with assistance from the Program owners/authors during February and March 1991. The plan was reviewed by SNL Management in April 1991. This document serves as a planning instrument for the Laboratories to aid in the scoping and sizing of activities related to ES H compliance for the coming five years. It will be modified as required to ensure a workload/funding balance and to address the findings resulting from the Tiger Team assessment at SNL, Albuquerque. The process of producing this document has served well to prepare SNL, Albuquerque, for the coming task of producing the required post-Tiger Team action plan document. 8 tabs.

The degree of oxidation of drops of aluminum metal was investigated parametrically for a range of melt diameters, relative melt-water velocities, melt temperatures, water temperatures, and ambient pressures using a combined film boiling-metal oxidation model. The model predictions of degree of oxidation were then compared to small-scale experiments involving molten drops of aluminum metal. The conclusions were, first, that for the range of melt temperatures and diameters considered, if an oxide layer forms and blankets the molten aluminum, then no significant oxidation occurs, in agreement with the results of experiments performed under quiescent, steady-state conditions. Second, comparing model results to data from single drop fragmentation experiments indicates that under the transient conditions occurring during rapid fragmentation, the oxide layer is disturbed and oxidation rates are limited primarily by the amount of steam available at the melt surface. Third, for a range of parameters, the heat gain in the melt drop from the oxidation reaction can exceed the heat loss to the surroundings, resulting in thermal runaway and ignition of the melt. This effect is observed experimentally as a threshold temperature effect, predicted to be about 1100 K for the initial single drop study, and between 1500 K and 1600 K for the single drop experiments. 7 refs., 2 tabs.

We present a method to conservatively estimate MOS hardness in space that shares the same technical basis as MIL-STD 883C, Test Method 1019.4, but permits greater latitude in part selection for low-dose space systems. Cobalt-60 irradiation at 50–300 rad(Si)/s followed by 25°C anneal is shown to provide an effective test of oxide-charge related failures at low dose rates that is considerably less conservative than Method 1019.4. For MOS devices with gate oxides thinner than 100 nm, we show that an elevated temperature “rebound test” generally is not required for systems with total dose requirements less than 5 krad(Si). For thicker gate oxides and/or higher-dose system requirements, rebound testing per Method 1019.4 generally is required to ensure that devices do not fail in space due to interface-trap effects. © 1991 IEEE

It is shown that preirradiation channel-mobility and 1/f-noise measurements can reveal a great deal about postirradiation interface- and oxide-trap charge buildup in MOS transistors. A model is developed to explain the observed correlations. 10 refs., 4 figs.

The effectiveness of hardening the SA3300 against SEU using design and design-plus-resistor approaches is presented. The full performance and SEU tolerance requirements are met using R{sub fb} {le} 160 k{Omega}. Lumped-parameter circuit simulations are used to analyze results. 10 refs., 5 figs.

Changes in brine and gas permeability of rock salt as a result of nearby excavation (mine-by) have been measured from the underground workings of the WIPP facility. Prior to the mine-by, the formation responds as a porous medium with a very low brine permeability, a significant pore (brine) pressure and no measurable gas permeability. The mine-by excavation creates a dilated, partially saturated zone in the immediate vicinity of the excavation with an increased permeability to brine and a measurable permeability to gas. The changes in hydrologic properties are discussed in the context of pore structure changes.

The chemical synthesis of advanced ceramic and glass materials by the sol-gel process has become an area of increasing activity in the field of material science. The sol-gel process provides a means to prepare homogeneous, high purity materials with tailored chemical and physical properties. This paper surveyed the nuclear magnetic resonance (NMR) studies of silicon-based sol-gel kinetics. A review of the various models which have been used to analyze the chemical kinetics of various sol-gel systems was presented. The utility of NMR spectroscopy was demonstrated in investigating the influence that various reaction conditions have on the reaction pathways by which sol-gel derived materials are synthesized. By observing in a direct fashion the chemical pathway of the sol-gel, it is often possible to relate the final properties of the material to the formulation and reaction conditions of the sol-gel. The study of reaction kinetics by NMR is expected to play an increasingly important role in understanding sol-gel processing and material properties. 15 refs. (DP)