Publications

The authors present sufficient conditions for the design of strictly positive real (SPR), fixed-order dynamic compensators. The primary motivation for designing SPR compensators is for application to positive real (PR) plants. When an SPR compensator is connected to a PR plant in a negative feedback configuration, the closed loop is guaranteed stable for arbitrary plant variations as long as the plant remains PR. Equations that are a modified form of the optimal projection equations, with the separation principle not holding in either the full- or reduced-order case, are given. A solution to the design equations in shown to exist when the plant is PR (or just stable). Finally, the closed-loop system consisting of a PR plant and an SPR compensator is shown to be S-structured Lyapunov stable.

A Nuisance Alarm Data System (NADS) was developed to gather long-term background alarm data on exterior intrusion detectors as part of their evaluation. Since nuisance alarms play an important part in the selection of intrusion detectors for use at Department of Energy (DOE) facilities, an economical and reliable way to monitor and record these alarms was needed. NADS consists of an IBM Personal Computer and printer along with other commercial units to communicate with the detectors, to gather weather data and to record video for assessment. Each alarm, its assessment and the weather conditions occurring at alarm time are placed into a database that is used in the evaluation of the detector. The operating software is written in Turbo Pascal for easy maintenance and modification. A portable system, based on the NADS design, has been built and shipped to other DOE locations to do on-site alarm monitoring. This has been valuable for the comparison of different detectors in the on-site environment and for testing new detectors when the appropriate conditions do not exist or cannot be simulated at the Exterior Intrusion Detection Testbed.

Statistical process control (SPC) of technology parameters relevant to radiation hardness, test structure to IC correlation, and extrapolation from laboratory to threat scenarios are keys to implementing QML for radiation hardness assurance in a cost-effective manner. Data from approximately 300 wafer lots fabricated in Sandia's 4/3-µm and CMOS IIIA (2-µm) technologies are used to demonstrate approaches to, and highlight issues associated with, implementing QML for radiation-hardened CMOS in space applications. An approach is demonstrated to implement QML for single-event upset (SEU) immunity on 16k SRAMs that involves relating values of feedback resistance to system error rates. It is seen that the process capability indices, Cp and Cpk, for the manufacture of 400 kΩ feedback resistors required to provide SEU tolerance do not conform to “6σ” quality standards. For total-dose, ΔVit shifts measured on transistors are correlated with circuit response in the space environment. SPC is illustrated for ΔVit, and violations of SPC rules are interpreted in terms of continuous improvement. Finally, design validation for SEU, and quality conformance inspections for total-dose, are identified as major obstacles to cost-effective QML implementation. Techniques and tools that will help QML provide real cost savings are identified as physical models, 3D device-plus-circuit codes, and improved design simulators. © 1990 IEEE

We have measured 1/f noise in MOS transistors as a function of gate and drain bias, total ionizing dose, and postirradiation biased annealing time. The transistors tested varied in size, radiation hardness, and process technology. The radiation-induced 1/f noise correlates strikingly with the oxide trap charge through irradiation and anneal, but not with interface-trap charge, for frequencies up to 10 kHz. This implies that oxide trapped charge is the pre-dominant factor which leads to the increased 1/f noise in irradiated MOS devices. © 1990 IEEE

A 2D-Laser Radar Imaging System consisting of a prototype 2D-Laser Radar Sensor and an Image Processing System is currently being developed as an intrusion detection system capable of immediate detection and quick assessment for perimeter security and surveillance. The objective of this system is to produce a thin laser wall as an invisible intrusion barrier. Since only a small space is needed to create a narrow laser radar wall, this system will work well where there is only limited or narrow zones available to create the secure perimeter. Images are created of objects which penetrate the laser radar wall for assessment and to determine the appropriate alarm response. Such a system can be used to protect against airborne threats from rooftop areas or to guard against ground threats across perimeter zones of critical facilities. This paper will discuss the operational concepts, the technology, and an initial performance of this prototype system. 2 refs., 6 figs.

The SEU vulnerability of the SA3300 16-bit microprocessor has been characterized, and the effects of two different design revisions on error rate have been explored. We found that the threshold for upset depends on the data pattern written into the general purpose registers. With all bits in the general purpose registers set to logic one, a design with 2-µm n- and p-channel transistor lengths had a threshold LET of 35 MeV-cm2/mg at 25°C and 4.5 volt operation. With all zero's stored in the registers the upset threshold increased by more than a factor of two to 83 MeV-cm2/mg. A second design revision, with 1.25-µm and 1.75-µm n- and p-channel transistor lengths, respectively, was more vulnerable to upset, but exhibited a smaller dependence on logic state. Measured threshold LET was 23 and 35 MeV-cm2/mg with all one's and all zero's, respectively. Microprobe measurements using a pulsed Nd:YAG laser suggest that the observed pattern dependence for both design revisions is due to bipolar photocurrent in a vertical n+pn transistor. A slight temperature dependence was observed in both design revisions. This is consistent with the use of oversized restoring transistors to minimize SEU vulnerability rather than polysilicon feedback resistors. More recent data show thresholds above 120 MeV-cm2/mg with 80 kΩ feedback resistors. © 1990 IEEE

The paper reviews the interaction of plasmas with materials and presents a status summary based on experience in large fusion experiments, laboratory investigations and design studies. The phenomena that are discussed limit the power densities and confinement properties attainable in plasmas; their control is an essential element of the design of future fusion devices. © 1990 IOP Publishing Ltd.

The use of porous oxide coatings, formed using sol-gel chemistry routes, as the discriminating elements of acoustic wave (AW) chemical sensors, is investigated. These coatings provide several unique advantages: durability, high adsorption capacity based on large surface areas, and chemical selectivity based on both molecular size and chemical interactions. The porosity of these coatings is determined by performing nitrogen adsorption isotherms using the AW device response to monitor the uptake of nitrogen at 77 K. The chemical sensitivity and selectivity obtained with this class of coatings is demonstrated using several examples: hydrous titanate ion exchange coatings, zeolite/silicate microcomposite coatings, and surface modified silicate films.

In light sources such as tungsten filament bulbs, fluorescent tubes and gas tube type radioluminescent (RL) lamps, visible light is emitted from a thin surface layer of excited material. In contrast, neon bulbs, xenon flash tubes and lasers emit light generated throughout their volumes. The first group can be characterized as surface emitters and the latter as volumetric emitters. Theoretically, an ideal volumetric light source has definite advantages over a surface source. In reality, practical volumetric sources will have limitations as well. These advantages and limitations will be discussed with particular emphasis on comparisons between current gas tube type RL lamps and the more advanced volumetric RL lamps.

A new generation of digital multimeters was used to compare the ratios of the resistances of wire-wound reference resistors and quantized Hall resistances. The accuracies are better than 0.1 ppM for ratios as large as 4:1 if the multimeters are calibrated with a Josephson array. 9 refs.

A class of borazene polymers was developed which consists of a two‐dimensional array of six‐membered borazene rings with the borons of adjacent borazene rings separated by ‐NH‐ groups. Pyrolysis of these polymers above ∼1000°C leads to crystalline graphite‐like boron nitride (h‐BN). The thermal chemistry of thin films of one polymer deposited on KOH‐eched aluminum was examined by thermal decomposition mass spectroscopy (TDMS) and thermal gravimetric analysis (TGA), and the gas evolution chemistry was found to be essentially complete at temperatures less than 400°C. All products desorb with the same temperature profile and the major desorbing species are NH3 and N2, consistent with a loss of excess nitrogen and hydrogen in the polymer, and HCl from decomposition of byproducts of the synthesis step. Since the formation of ordered crystalline h‐BN films requires heating to temperatures of the order of 1000°C, whereas the gas evolution chemistry is complete by roughly 400°C, it is concluded that gas evolution chemical processes are not rate limiting in BN ceramic production. Copyright © 1990, Wiley Blackwell. All rights reserved

Constitutive modeling and bifurcation analyses are combined with axisymmetric (triaxial), triaxial/torsion, and plane-strain experiments to interpret and anticipate the development of shear localization in rocks. This paper discusses preliminary results. 22 refs., 3 figs.

The US Department of Energy (DOE) has supported the development of the sodium-sulfur technology since 1973. The programs have focused on progressing core aspects of the technology and completing initial battery engineering for both mobile and stationary applications. An overview of the Office of Energy Management (OEM) activities is contained in this paper. Two major development programs have been active: the first with Ford Aerospace and Communications Corporation (1975 to 1985), and the second with Chloride Silent Power Limited (1985 to 1990). With the completion this year of the qualification of a cell suitable for initial Solar Energy Systems (SES) applications, the emphasis of future DOE/OEM sodium/sulfur programs will shift to SES-battery engineering and development. The initial effort will resolve a number of issues related to the feasibility of utilizing the sodium/sulfur technology in these large-scale applications. This multi-year activity will represent the initial phase of an integrated long-term DOE-supported program to produce a commercially viable battery system.

The integrity of many mechanical assemblies and electrical components depends on small threaded fasteners. The design standards for small (less than {1/4} inch in diameter) screws made of stainless steel are not as well developed as those for larger sizes of high strength steels. The typical design approach is based on the application of static design principals. Steady state accelerations are applied to the component or assembly and sufficient screws are installed in mounting hardware for attachment to the next assembly. These design principals have been used successfully for years in a wide variety of applications. As the parts requiring small screws have continued to decrease in size, some design requirements include greater thread depths and adherence to strict interpretation of the governing thread standards. These design requirements have their origins in the lack of adequate definitions and standards for designs using small threaded fasteners. These design practices have led to significant problems in manufacturing parts with small threaded fasteners by requiring thread depths to four and more diameters of engagement while maintaining thread heights (radial engagement) of 75 percent throughout the thread interfaces. A test program was developed to address questions regarding design and manufacturing issues involving small threaded fasteners which included tensile strength, length of engagement needed to achieve the full strengths of the screws, and verification of the static design principals in dynamic conditions. This paper summarizes the initial results obtained to date from this test program and describes the work-in-progress on the dynamic tests with their related static tests. 2 refs., 4 figs., 4 tabs.

An isotropic continuum damage theory which accounts for the degradation of material strength under quasi-static loading conditions has been developed in the present investigation. The damage mechanism in this theory has been selected to be the interaction and growth of subscale cracks. The development of the theory follows closely the strain-rate dependent dynamic model advanced by the first author and his coworkers. Briefly, the cracks are activated by the maximum principal tensile strain and the density of activated cracks is described by a Weibull statistical distribution. The moduli of a cracked solid derived by Budiansky and O'Connell are then used to represent the global material degradation due to subscale cracking. Two additional material constants have been introduced in this model. These constants are determined from uniaxial tensile test data. The model has been implemented into a finite element code. Sample calculations involving the uniaxial and biaxial responses of plain concrete panels are presented to demonstrate the utility of the model. 7 refs., 2 figs.

The first in a series of multi-factor experiments designed to optimize the chemical cleaning procedure for four types of silicon material used in solar cell fabrication has been completed. The goal of this first experiment (a twenty-two factor main-effects experiment) was to determine the factors associated with chemical cleaning procedures that are most important in obtaining high excess charge-carrier recombination lifetime following a high-temperature furnace oxidation. It was determined that the factors having the strongest influence on charge-carrier lifetime were different for the four different silicon materials considered. In general, the lower the lifetime of the material, the less sensitive the material was to different chemical cleaning steps. The stability of the lifetime was also evaluated with several factors exhibiting a significant effect for high-quality silicon. Chemical cleaning procedures were identified that resulted in stable post-oxidation lifetimes greater than 2 ms for high-resistivity float-zone silicon. 3 refs., 8 figs.

The closure measurements from a large scale, heated, in situ experimental room in salt are compared to numerical calculations using the most recent predictive technology, with very good agreement, limited potentially only by the unmodeled roof fracture and separation.

Insulations are used in metallic glass ribbon cores in pulse power applications to prevent interlaminar eddy currents due to voltages induced between adjacent laminations. These interlaminar eddy currents can greatly increase the losses in cores, and, thereby, decrease the pulse permeability at high magnetization rates. This paper reports results of experiments with various insulation materials and both low and high induced anisotropy energy iron-base metallic glass ribbons. Co-wound insulation films as well as conformal insulations were investigated. Magnetic properties and voltage hold-off strengths are reported. 11 refs., 11 figs., 5 tabs.

When quartz controlled oscillators are required for use in applications demanding precision many factors will ultimately place limitations on the ability of the oscillator to remain at the desired frequency. These factors include temperature, resonator Q, pullability, radiation, output load variability, and the electronic components. This paper addresses the subject of frequency instability of oscillator circuits due to power supply voltage variations. In particular, the primary sources of this instability are described for a Pierce oscillator employing a bipolar transistor and design techniques are presented which minimize these frequency pulling effects. 4 refs., 17 figs.

The three-dimensional structural analysis of reentry vehicles presents a considerable challenge to the analyst. This is due to the mechanics of the problem as well as the incorporation of results from several disciplines into the environmental description of the problem. Separate results from aero-analyses, frequently computed in one-dimensional format must be combined into a three-dimensional format suitable for a structural finite element analysis. Features required for the analysis include the ablated thickness of the heatshield structure, as well as pressure on the vehicle and the temperature distribution through the heatshield. By combining these environments, a complete description of all factors which affect the structural performance of reentry vehicles are included into one analysis. This paper presents a method of analyzing the structural response of reentry vehicles using the complete three-dimensional environmental load description.

Reflection Mass Spectrometry (REMS) consists of a cryo-shrouded mass spectrometer which measures mass-analyzed, line-of-sight chemical fluxes from a growing wafer. It is especially useful during III/V molecular beam epitaxy (MBE) for which there are always substantial group V fluxes and often some group III fluxes leaving the wafer during growth. These fluxes depend sensitively on the instantaneous chemical reactivity of the surface. That chemical reactivity in turn depends on instantaneous alloy composition (III/III ratio), surface stoichiometry (As coverage) and temperature. In this brief summary of our work, we describe two examples of the engineering'' usefulness of REMS, involving MBE of InAlAs and InGaAs and one example of measurements of basic scientific interest. 3 figs.

A program is under way to develop liquid metal heat pipes that can transfer energy from the focal point of a parabolic solar concentrator to the heater tubes of one or more Stirling engines. To design high performance wicks for heat pipe solar receivers, it is necessary to have an accurate assessment of the wick's properties. Procedures for measuring the flow properties of wicks before and after fabrication processes take place are presented. The testing procedures provide a useful method of determining the validity of a wick design before full-scale testing is attempted.

Significant progress is continuing in the development of photovoltaic (PV) concentrator technology. New record cell and module efficiencies have been achieved, and improvements in cells, cell assemblies, and modules are increasing reliability and decreasing cost. The number of firms actively pursuing PV concentrator module technology has increased substantially in the last three years. Two new concentrator systems were installed last year, and we are likely to see more in the near future. This paper describes the most significant developments of the last two years, including descriptions of PV concentrator module development and reliability activities, advances in concentrator cell technology, the new PV concentrator array installations, a new Concentrator Initiative Program, and results of the latest costing study. 26 refs., 9 figs., 1 tab.

Chacahoula salt dome, eight miles southwest of Thibodaux, LA, could be solution mined to create caverns for storing as much as 500 million barrels (MMB) of crude oil, should the Strategic Petroleum Reserve (SPR) require additional storage volume. The salt mass geometry is confirmed by more than 50 oil wells, and also from previous exploratory drilling for sulphur. Top of salt occurs at {minus}1100 ft, and some 1300 acres exist within the {minus}2000 ft salt contour. Frasch mining of 1.35 million long tons of sulphur caused the surface to subside about one foot on the northeastern part of the dome. Creep-induced subsidence averaging {approximately}2.7 ft over 30 yrs is estimated for a 200 MMB cavern array, which would require perimeter diking to control localized perennial flooding. Earthquakes approaching intensity MM 6 have occurred nearby and are expected to recur on the order of {approximately}100 yrs but would not affect cavern stability. Additional study of brine disposal methods and hurricane surge probabilities are needed to establish design parameters and cost estimates for storage. 11 refs., 8 figs., 2 tabs.

A Nuisance Alarm Data System (NADS) was developed to gather long- term background alarm data on exterior intrusion detectors as part of their evaluation. Since nuisance alarms play an important part in the selection of intrusion detectors for use at Department of Energy (DOE) facilities, an economical and reliable way to monitor and record there alarms was needed. NADS consists of an IBM Personal Computer and printer along with other commercial units to communicate with detectors, to gather weather data and to record video for assessment. Each alarm, its assessment and the weather conditions occurring at alarm time are placed into a database that is used in the evaluation of the detector. The operating software is written in Turbo Pascal for easy maintenance and modification. A portable system, based on the NADS design, has been built and shipped to other DOE locations to do on-site alarm monitoring. This has been valuable for the comparison of different detectors in the on-site environment and for testing new detectors when the appropriate conditions do not exist or cannot be simulated at the Exterior Intrusion Detection Testbed.