Publications

Sub-optimal design practices can reduce the radiation hardness of a circuit even though it is fabricated in a radiation hardened process. This is especially true for a nonvolatile memory, as compared to a standard digital circuit, where high voltages and unusual bias conditions are required. This paper will discuss the design techniques used in the development of a 64K EEPROM (Electrically Erasable Programmable Read Only Memory) to maximize radiation hardness. The circuit radiation test results will be reviewed in order to provide validation of the techniques. © 1993 IEEE

The U.S. Department of Energy (DOE) is developing the Waste Isolation Pilot Plant (WIPP) in southeastern New Mexico as a facility for the long-term disposal of defense-related transuranic (TRU) wastes. Use of the WIPP for waste disposal is contingent on demonstrations of compliance with applicable regulations of the U.S. Environmental Protection Agency (EPA). This paper addresses issues related to modeling gas and brine migration at the WIPP for compliance with both EPA 40 CFR 191 (the Standard) and 40 CFR 268.6 (the RCRA). At the request of the WIPP Project Integration Office (WPIO) of the DOE, the WIPP Performance Assessment (PA) Department of Sandia National Laboratories (SNL) has completed preliminary uncertainty and sensitivity analyses of gas and brine migration away from the undisturbed repository. This paper contains descriptions of the numerical model and simulations, including model geometries and parameter values, and a summary of major conclusions from sensitivity analyses. Because significant transport of contaminants can only occur in a fluid (gas or brine) medium, two-phase flow modeling can provide an estimate of the distance to which contaminants can migrate. Migration of gas or brine beyond the RCRA 'disposal-unit boundary' or the Standard's accessible environment constitutes a potential, but not certain, violation and may require additional evaluations of contaminant concentrations.

This paper presents an infinite impulse response (IIR) filtering technique for reducing structural vibration in remotely operated robotic systems. The technique uses a discrete filter between the operator's joy stick and the robot controller to alter the inputs of the system so that residual vibration and swing are reduced. A linearized plant model of the system is analyzed in the discrete time domain, and the filter is designed using pole-zero placement in the z-plane. This technique has been successfully applied to a two link flexible arm and a gantry crane with a suspended payload.

Before disposing of transuranic radioactive waste at the Waste Isolation Pilot Plant (WIPP), the United States Department of Energy (DOE) must evaluate compliance with long-term regulations of the United States Environmental Protection Agency (EPA). Sandia National Laboratories (SNL) is conducting iterative performance assessments (PAs) of the WIPP for the DOE to provide interim guidance while preparing for final compliance evaluations. This paper describes the 1992 preliminary comparison with Subpart B of the Environmental Standards for the Management and Disposal of Spent Nuclear Fuel, High-Level and Transuranic Radioactive Wastes (40 CFR 191), which regulates long-term releases of radioactive waste. Results of the 1992 PA are preliminary, and cannot be used to determine compliance or noncompliance with EPA regulations because portions of the modeling system and data base are incomplete. Results are consistent, however, with those of previous iterations of PA, and the SNL WIPP PA Department has high confidence that compliance with 40 CFR 191B can be demonstrated. Comparison of predicted radiation doses from the disposal system also gives high confidence that the disposal system is safe for long-term isolation.

Before disposing of transuranic radioactive waste at the Waste Isolation Pilot Plant (WIPP), the United States Department of Energy (DOE) must evaluate compliance with long-term regulations of the United States Environmental Protection Agency (EPA), specifically the Environmental Standards for the Management and Disposal of Spent Nuclear Fuel, High-Level and Transuranic Radioactive Wastes (40 CFR 191), and the Land Disposal Restrictions (40 CFR 268) of the Hazardous and Solid Waste Amendments to the Resource Conservation and Recovery Act (RCRA). Sandia National Laboratories (SNL) is conducting iterative performance assessments (PAs) of the WIPP for the DOE to provide interim guidance while preparing for final compliance evaluations. This paper provides background information on the regulations, describes the SNL WIPP PA Department's approach to developing a defensible technical basis for consistent compliance evaluations, and summarizes the major observations and conclusions drawn from the 1991 and 1992 PAs.

Two space applications are considered for (electrically) contactless coilguns: launch of small satellites into low-earth orbit, and launch of lunar liquid oxygen (LLOX) from the moon to the stationary Lagrangian point L2. For the earth-to-orbit (ETO) application, the baseline conceptual design consists of a 960-m long gun sited in a tunnel at 25° inclination. The gun launches an 1820-kg package that includes a 100-kg satellite and a 650-kg boost rocket for orbital insertion. For the lunar application, the launcher is 200 m long. A 100-kg load of LLOX is packaged in a 10-kg fiber-wrapped tank, accelerated at 2 kgees in an aluminum bucket (armature), and launched at 2.33 km/s at 30-minute intervals. The canisters arrive at L2 2.97 days later and are captured by robotic tugs that deliver them to a fuel depot. The total mass of LLOX delivered per year is 867 Mg (metric tons). © 1993 IEEE

We postulate that there is a required velocity of magnetic flux into a plasma-armature railgun to keep the armature from expanding toward the breech and thereby changing the nature of the forward-going force. Testing this idea with data from several guns we find confirmation, provided the required flux velocity increases with the square of the linear current density. This leads to the conclusion that these guns have a momentum limit while they are operating as true electromagnetic launchers. That limit is derived. High-performance railguns reach this limit during current rise, and most of their momentum comes from a hybrid phase. The hybrid phase has a pressure limitation at high velocity similar to electrothermal, light-gas, and powder guns. The 6 km/s “velocity limit” is explained by this pressure limitation. © 1993 IEEE

Different hardness-assurance tests are often required for advanced bipolar devices than for CMOS devices. In this work, the dose-rate dependence of bipolar current-gain degradation is mapped over a wide range of dose rates for the first time, and it is very different from analogous MOSFET curves. Annealing experiments following irradiation show negligible change in base current at room temperature, but significant recovery at temperatures of 100°C and above. In contrast to what is observed in MOSFET’s, irradiation and annealing tests cannot be used to predict the low-dose-rate response of bipolar devices. A comparison of x-ray-induced and 60Co gamma-ray-induced gain degradation is reported for the first time for bipolar transistors. The role of the emitter bias during irradiation is also examined. Implications for hardening and hardness assurance are discussed. © 1993 IEEE

Vadose-zone moisture transport near an impermeable barrier has been under study at a field site near Albuquerque, NM since 1990. Moisture content and temperature have been monitored in the subsurface on a regular basis; both undergo a seasonal variation about average values. Even though the slab introduces two-dimensional effects on the scale of the slab, moisture and heat transport is predominantly vertical. Numerical simulations, based on the models developed by Philip and de Vries (1957) and de Vries (1958), indicate that the heat flow is conduction-dominated while the moisture movement is dominated by diffusive vapor distillation. Model predictions of the magnitude and extent of changes in moisture content underneath the slab are in reasonable agreement with observation.

A Sandia National Laboratories/AT&T Bell Laboratories Team is developing a soft x-ray projection lithography tool that uses a compact laser plasma as a source of 14 nm x-rays. Optimization of the 14 nm x-rays source brightness is a key issue in this research. This paper describes our understanding of the source as it has been obtained through the use of computer simulations utilizing the LASNEX radiation-hydrodynamics code.

Ion-beam-induced charge-collection imaging (IBICC) has been used to study the SEU mechanisms of the Sandia TA670 16K-bit SRAM. Quantitative charge-collection spectra from known regions of the memory cell have been derived with this technique. For 2.4-MeV He ions at normal incidence, charge collection depth for a reverse-biased p+ drain strike is estimated to be 4.8±0.4 μm. Heavy-ion strikes to the reverse-biased p-well result in nearly complete collection of deposited charge to a depth of 5.5±0.5 μm. A charge amplification effect in the n-on drain is identified and is due to either bipolar amplification or a shunt effect in the parasitic vertical npn bipolar transistor associated with the n+/n substrate, p-well, and n+ drain. This effect is present only when the n+ drain is at 0V bias. When coupled with previous SEU-imaging, these results strongly suggest that the dominant SEU mechanism in this SRAM is a heavy-ion strike to the n-on transistor drain. © 1993 IEEE

The Vapnik-Chervonenkis (V-C) dimension is an important combinatorial tool in the analysis of learning problems in the PAC framework. For polynomial learnability, we seek upper bounds on the V-C dimension that are polynomial in the syntactic complexity of concepts. Such upper bounds are automatic for discrete concept classes, but hitherto little has been known about what general conditions guarantee polynomial bounds on V-C dimension for classes in which concepts and examples are represented by tuples of real numbers. In this paper, we show that for two general kinds of concept class the V-C dimension is polynomially bounded as a function of the syntactic complexity of concepts. One is classes where the criterion for membership of an instance in a concept can be expressed as a formula (in the first-order theory of the reals) with fixed quantification depth and exponentially-bounded length, whose atomic predicates are polynomial inequalities of exponentially-bounded degree. The other is classes where containment of an instance in a concept is testable in polynomial time, assuming we may compute standard arithmetic operations on reals exactly in constant time. Our results show that in the continuous case, as in the discrete, the real barrier to efficient learning in the Occam sense is complexity-theoretic and not information-theoretic. We present examples to show how these results apply to concept classes defined by geometrical figures and neural nets, and derive polynomial bounds on the V-C dimension for these classes.

During hydrocarbon reservoir stimulations, such as hydraulic fracturing, the cracking and slippage of the formation results in the emission of seismic energy. The objective of this study was to determine the properties of these induced micro-seisms. A hydraulic fracture experiment was performed in the Piceance Basin of Western Colorado to induce and record micro-seismic events. The formation was subjected to four processes; breakdown/ballout, step-rate test, KCL mini-fracture, and linear-gel mini-fracture. Micro-seisms were acquired with an advanced three-component wall-locked seismic accelerometer package, placed in an observation well 211 ft offset from the fracture well. During the two hours of formation treatment, more than 1200 micro-seisms with signal-to-noise ratios in excess of 20 dB were observed. The observed micro- seisms had a nominally flat frequency spectrum from 100 Hz to 1500 Hz and lack the spurious tool-resonance effects evident in previous attempts to measure micro-seisms. Both p-wave and s-wave arrivals are clearly evident in the data set, and hodogram analysis yielded coherent estimates of the event locations. This paper describes the characteristics of the observed micro- seismic events (event occurrence, signal-to-noise ratios, and bandwidth) and illustrates that the new acquisition approach results in enhanced detectability and event location resolution.

Many designs use EPLDs (Erasable Programmable Logic Devices) to implement control logic and state machines. If the design is slow, timing through the EPLD is not crucial so designers often treat the device as a black box. In high speed designs, timing through the EPLD is critical. In these cases a thorough understanding of the device architecture is necessary. Lessons learned in the implementation of a high-speed design using the Altera EPM5130 are discussed.

A new Assembly Test Chip, ATC04, designed to measure mechanical stresses at the die surface has been built and tested. This CMOS chip 0.25 in. on a side, has an array of 25 piezoresistive stress sensing cells, four resistive heaters and two ring oscillators. The ATCO4 chip facilitates making stress measurements with relatively simple test equipment and data analysis. The design, use, and accuracy of the chip are discussed and initial results are presented from three types of stress measurement experiments: four-point bending calibration, single point bending of a substrate with an ATC04 attached by epoxy, and stress produced by a liquid epoxy encapsulant.

The feasibility of utilizing a groundbased laser without an orbital mirror for space debris removal is examined. Technical issues include atmospheric transmission losses, adaptive-optics corrections of wavefront distortions, laser field-of-view limitations, and laser-induced impulse generation. The physical constraints require a laser with megawatt output, long run-time capability, and wavelength with good atmospheric transmission characteristics. It is found that a 5-MW reactor-pumped laser can deorbit debris having masses of the order of one kilogram from orbital altitudes to be used by Space Station Freedom. Debris under one kilogram can be deorbited after one pass over the laser site, while larger debris can be deorbited or transferred to alternate orbits after multiple passes over the site.

Flux-flow based devices such as the superconducting flux flow transistor and magnetically controlled long junctions have been made from thin films of TICaBaCuO and YBaCuO. The devices are based on the magnetic control of flux flow in their respective structures: a long junction or an array of weak links. The equivalent circuits of the two devices are similar: a low impedance input control line, an output impedance of 3-20 Q and an active current-controlled element. The long junctions have tended to be slower, have lower gain and be somewhat less noisy than their counterparts. The performance of circuits such as narrowband and distributed amplifiers (SO GHz bandwidths, noise figures < 3 dB), phase shifters (continuous with < 2 dB loss 4-40 GHz), logic gates (2-3 ps gate delays) and memories made using these devices will be compared and analyzed. © 1993 IEEE

Automatic motion planning of a spray cleaning robot with collision avoidance is presented in this paper. In manufacturing environments, electronic and mechanical components are traditionally cleaned by spraying or dipping them using chlorofluorocarbon (CFC) solvents. As new scientific data show that such solvents are major causes for stratospheric ozone depletion, an alternate cleaning method is needed. Part cleaning with aqueous solvents is environmentally safe, but can require precision spraying at high pressures for extended time periods. Operator fatigue during manual spraying can decrease the quality of the cleaning process. By spraying with a robotic manipulator, the necessary spray accuracy and consistency to manufacture high-reliability components can be obtained. Our motion planner was developed to automatically generate motions for spraying robots based on the part geometry and cleaning process parameters. For spraying paint and other coatings a geometric description of the parts and robot may be sufficient for motion planning, since coatings are usually done over the visible surfaces. For spray cleaning, the requirement to reach hidden surfaces necessitates the addition of a rule-based method to the geometric motion planning.

The geochemical properties of a porous sand and several tracers (Ni, Br, and Li) have been characterized for use in a caisson experiment designed to validate sorption models used in models of reactive transport. The surfaces of the sand grains have been examined by a combination of techniques including potentiometric titration, acid leaching, optical microscopy, and scanning electron microscopy with energy-dispersive spectroscopy. The surface studies indicate the presence of small amounts of carbonate, kaolinite and iron-oxyhydroxides. Adsorption of nickel, lithium and bromide by the sand was measured using batch techniques. Bromide was not sorbed by the sand. A linear (Kd) or an isotherm sorption model may adequately describe transport of Li; however, a model describing the changes of pH and the concentrations of other solution species as a function of time and position within the caisson and the concomitant effects on Ni sorption may be required for accurate predictions of nickel transport.

For problems where media properties are measured at one scale and applied at another, scaling laws or models must be used in order to define effective properties at the scale of interest. The accuracy of such models will play a critical role in predicting flow and transport through the Yucca Mountain Test Site given the sensitivity of these calculations to the input property fields. Therefore, a research program has been established to gain a fundamental understanding of how properties scale with the aim of developing and testing models that describe scaling behavior in a quantitative manner. Scaling of constitutive rock properties is investigated through physical experimentation involving the collection of suites of gas permeability data measured over a range of discrete scales. Also, various physical characteristics of property heterogeneity and the means by which the heterogeneity is measured and described and systematically investigated to evaluate their influence on scaling behavior. This paper summarizes the approach that is being taken toward this goal and presents the results of a scoping study that was conducted to evaluate the feasibility of the proposed research.

Experimental results exploring gravity-driven wetting front instability in a pre-wetted, rough-walled analog fracture are presented. Initial conditions considered include a uniform moisture field wetted to field capacity of the analog fracture and the structured moisture field created by unstable infiltration into an initially dry fracture. As in previous studies performed under dry initial conditions, instability was found to result both at the cessation of stable infiltration and at flux lower than the fracture capacity under gravitational driving force. Individual fingers were faster, narrower, longer, and more numerous than observed under dry initial conditions. Wetting fronts were found to follow existing wetted structure, providing a mechanism for rapid recharge and transport.

In an attempt to achieve completeness and consistency, the performance-assessment analyses developed by the Yucca Mountain Project are tied to scenarios described in event trees. Development of scenarios requires describing the constituent features, events, and processes in detail. Several features and processes occurring at the waste packages and the rock immediately surrounding the packages (i.e., the near field) have been identified: the effects of radiation on fluids in the near-field rock, the path-dependency of rock-water interactions, and the partitioning of contaminant transport between colloids and solutes. This paper discusses some questions regarding these processes that the near-field performance-assessment modelers will need to have answered to specify those portions of scenarios dealing with the near field.

Experiments investigating the behavior of individual, gravity-driven fingers in an initially dry, rough-walled analog fracture are presented. Fingers were initiated from constant flow to a point source. Finger structure is described in detail; specific phenomena observed include: desaturation behind the finger-tip, variation in finger path, intermittent flow structures, finger-tip bifurcation, and formation of dendritic sub-fingers. Measurements were made of finger-tip velocity, finger width, and finger-tip length. Non-dimensional forms of the measured variables are analyzed relative to the independent parameters, flow rate and gravitational gradient.

The potential repository system is intended to isolate high-level radioactive waste at Yucca Mountain according to the performance objective - 10 CFR 60.112. One subsystem that may contribute to achieving this objective is the sealing subsystem. This subsystem is comprised of sealing components in the shafts, ramps, underground network of drifts, and the exploratory boreholes. Sealing components can be rigid, as in the case of a shaft seal, or can be more compressible, as in the case of drift fill comprised of mined rockfill. This paper presents the preliminary seismic response of discrete sealing components in welded and nonwelded tuff. Special consideration is given to evaluating the stress in the seal, and the behavior of the interface between the seal and the rock. The seismic responses are computed using both static and dynamic analyses. Also presented is an evaluation of the maximum seismic response encountered by a drift seal with respect to the angle of incidence of the seismic wave. Mitigation strategies and seismic design considerations are proposed which can potentially enhance the overall response of the sealing component and subsequently, the performance of the overall repository system.

This analytical study examines the effect of initial thermodynamic conditions on the loads generated by the combustion of homogeneous hydrogen-air-steam mixtures. The effect of initial temperature, pressure, hydrogen concentration, and steam concentration is evaluated for two cases, (1) constant volume and (2) constant initial pressure. For each case, the Adiabatic, Isochoric, Complete Combustion (AICC), Chapman-Jouguet (CJ), and normally reflected CJ pressures are calculated for a range of hydrogen and steam concentrations representative of the entire flammable regime. For detonation loads, pressure profiles and time-histories are also evaluated in one-dimensional Cartesian geometry. The results show that to a first approximation, the AICC and CJ pressures are directly proportional to the initial density. Increasing the hydrogen concentration up to stoichiometric concentrations significantly increases the AICC, CJ, and reflected CJ pressures. For the constant volume case, the AICC, CJ, and reflected CJ pressures increase with increasing hydrogen concentration on the rich side of stoichiometric concentrations. For the constant initial pressure case, the AICC, CJ, and reflected CJ pressures decrease with increasing hydrogen concentration on the rich side of stoichiometric values. The addition of steam decreases the AICC, CJ, and reflected CJ pressures for the constant initial pressure case, but increases them for the constant volume case. For detonations, the pressure time-histories can be normalized with the AICC pressure and the reverberation time for Cartesian geometry. © 1993.