Publications

Etch rates up to 7000{angstrom}/min for InP and 3500{angstrom}/min for GaAs are obtained for high microwave power (1000W) CH{sub 4}/H{sub 2}/Ar Electron Cyclotron Resonance plasma etching. Preferential loss of the group V element leads to nonstoichiometric, unacceptably rough surfaces on In-based binary semiconductors at microwave powers {ge}400W, regardless of plasma composition. Both Ga- and Al-based materials retain smooth, stoichiometric surfaces even at I000W, but the rates are still much slower than for C1{sub 2} plasma chemistries. The results suggest that CH{sub 4}/H{sub 2} plasmas are not well suited to ECR systems operating at high powers.

Structural system simulation is important in analysis, design, testing, control, and other areas, but it is particularly difficult when the system under consideration is nonlinear. Artificial neural networks offer a useful tool for the modeling of nonlinear systems, however, such modeling may be inefficient or insufficiently accurate when the system under consideration is complex. This paper shows that there are several transformations that can be used to uncouple and simplify the components of motion of a complex nonlinear system, thereby making its modeling and simulation a much simpler problem. A numerical example is also presented.

Short communication.

The Geothermal Drilling Organization (GDO), formed in the early 1980s by the geothermal industry and the U.S. Department of Energy (DOE) Geothermal Division, sponsors specific development projects to advance the technologies used in geothermal exploration, drilling, and production phases. Individual GDO member companies can choose to participate in specific projects that are most beneficial to their industry segment. Sandia National Laboratories is the technical interface and contracting office for the DOE in these projects. Typical projects sponsored in the past have included a high temperature borehole televiewer, drill bits, muds/polymers, rotary head seals, and this project for drill pipe protectors. This report documents the development work of Regal International for high temperature geothermal pipe protectors.

Thermal and nonthermal x-ray emission from the implosion of compact tungsten wire arrays, driven by 5 MA from the Saturn accelerator, are measured and compared with LLNL Radiation-Hydro-Code (RHC) and SNL Hydro-Code (HC) numerical models. Multiple implosions, due to sequential compressions and expansions of the plasma, are inferred from the measured multiple x-radiation bursts. Timing of the multiple implosions and the thermal x-ray spectra measured between 1 and 10 keV are consistent with the RHC simulations. The magnitude of the nonthermal x-ray emission measured from 10 to 100 keV ranges from 0.02 to 0.08% of the total energy radiated and is correlated with bright-spot emission along the z-axis, as observed in earlier Gamble-11 single exploding-wire experiments. The similarities of the measured nonthermal spectrum and bright-spot emission with those measured at 0.8 MA on Gamble-II suggest a common production mechanism for this process. A model of electron acceleration across magnetic fields in highly-collisional, high-atomic-number plasmas is developed, which shows the existence of a critical electric field, E{sub c}, below which strong nonthermal electron creation (and the associated nonthermal x rays) do not occur. HC simulations show that significant nonthermal electrons are not expected in this experiment (as observed) because the calculated electric fields are at least one to two orders-of-magnitude below E{sub c}. These negative nonthermal results are confirmed by RHC simulations using a nonthermal model based on a Fokker-Plank analysis. Lastly, the lower production efficiency and the larger, more irregular pinch spots formed in this experiment relative to those measured on Gamble II suggest that implosion geometries are not as efficient as single exploding-wire geometries for warm x-ray production.

This is the summary report for the Protocol Extensions for Asynchronous Transfer Mode project, funded under Sandia`s Laboratory Directed Research and Development program. During this one-year effort, techniques were examined for integrating security enhancements within standard ATM protocols, and mechanisms were developed to validate these techniques and to provide a basic set of ATM security assurances. Based on our experience during this project, recommendations were presented to the ATM Forum (a world-wide consortium of ATM product developers, service providers, and users) to assist with the development of security-related enhancements to their ATM specifications. As a result of this project, Sandia has taken a leading role in the formation of the ATM Forum`s Security Working Group, and has gained valuable alliances and leading-edge experience with emerging ATM security technologies and protocols.

An investigation of the shock compression properties of high-strength ceramics has been performed using controlled planar impact techniques. In a typical experimental configuration, a ceramic target disc is held stationary, and it is struck by plates of either a similar ceramic or by plates of a well-characterized metal. All tests were performed using either a single-stage propellant gun or a two-stage light-gas gun. Particle velocity histories were measured with laser velocity interferometry (VISAR) at the interface between the back of the target ceramic and a calibrated VISAR window material. Peak impact stresses achieved in these experiments range from about 3 to 70 GPa. Ceramics tested under shock impact loading include: Al{sub 2}O{sub 3}, AlN, B{sub 4}C, SiC, Si{sub 3}N{sub 4}, TiB{sub 2}, WC and ZrO{sub 2}. This report compiles the VISAR wave profiles and experimental impact parameters within a database-useful for response model development, computational model validation studies, and independent assessment of the physics of dynamic deformation on high-strength, brittle solids.

This study provides a scoping safety assessment for disposal of toxic metals contained in Department of Energy (DOE) mixed low-level waste (MLLW) at six DOE sites that currently have low-level waste (LLW) disposal facilities--Savannah River Site, Oak Ridge Reservation, Los Alamos National Laboratory, Hanford Reservation, Nevada Test Site, and Idaho National Engineering Laboratory. The study has focused on the groundwater contaminant pathway, which is considered to be the dominant human exposure pathway from shallow land MLLW disposal. A simple and conservative transport analysis has been performed using site hydrological data to calculate site-specific ``permissible`` concentrations of toxic metals in grout-immobilized waste. These concentrations are calculated such that, when toxic metals are leached from the disposal facility by infiltrating water and attenuated in local ground-water system the toxic metal concentrations in groundwater below the disposal facility do not exceed the Maximum Contaminant Levels as stated in the National Primary Drinking Water Regulation. The analysis shows that and sites allow about I00 times higher toxic metal concentrations in stabilized waste leachate than humid sites. From the limited available data on toxic metal concentrations in DOE MLLW, a margin of protection appears to exist in most cases when stabilized wastes containing toxic metals are disposed of at the DOE sites under analysis. Possible exceptions to this conclusion are arsenic, chromium selenium, and mercury when disposed of at some humid sites such as the Oak Ridge Reservation. This analysis also demonstrates that the US Environmental Protection Agency`s prescriptive regulatory approach that defines rigid waste treatment standards does not inherently account for the variety of disposal environments encountered nationwide and may result in either underprotection of groundwater resources (at humid sites) or an excessive margin of protection (at and sites).

A brief overview is given for two software developments related to the ITS code system. These developments provide parallel processing and visualization capabilities and thus allow users to perform ITS calculations more efficiently. Timing results and a graphical example are presented to demonstrate these capabilities.

The ability to rapidly command multi-robot behavior is crucial for the acceptance and effective utilization of multiple robot control. To achieve this, a modular- multiple robot control solution is being, pursued using the SMART modular control architecture. This paper investigates the development of a new dual-arm kinematics module (DUAL-KLN) which allows multiple robots, previously controlled as separate stand-alone systems, to be controlled as a coordinated multi-robot system. The DUAL-KIN module maps velocity and force information from a center point of interest on a grasped object to the tool centers of each grasping robot. Three-port network equations are used and mapped into the scattering operator domain in a computationally efficient form. Application examples of the DUAL-KLN module in multi-robot coordinated control are given.

This paper analyses the problems and presents solutions for building a modular robot control system. The approach requires modeling the entire robot system using multi-dimensional passive networks, breaking the system into subnetwork ``modules,`` and then discretizing the subnetworks, or n-ports, in a passivity preserving fashion. The main difficulty is the existence of ``algebraic loops`` in the discretized system. This problem is overcome by the use of scattering theory, whereby the inputs and outputs of the n-ports are mapped into wave variables before being discretized. By first segmenting the n-ports into nonlinear memoryless subnetworks and linear dynamic subnetworks and then discretizing using passivity preserving techniques such as Tustin`s method, a complete modular robot control solution is obtained.

The characteristics of a piezoresistive accelerometer in shock environments are being studied at Sandia National Laboratories in the Mechanical Shock Testing Laboratory. A Hopkinson bar capability has been developed to extend our understanding of the piezoresistive accelerometer, in two mechanical configurations, in the high frequency, high shock environments where measurements are being made. In this paper, the beryllium Hopkinson bar configuration with a laser doppler vibrometer as the reference measurement is described. The in-axis performance of the piezoresistive accelerometer for frequencies of dc-50 kHz and shock magnitudes of up to 70,000 g as determined from measurements with a beryllium Hopkinson bar are presented. Preliminary results of characterizations of the accelerometers subjected to cross-axis shocks in a split beryllium Hopkinson bar configuration are presented.

An advanced packaging concept has been developed for optical devices. This concept allows multiple fibers to be coupled to photonic integrated circuits, with no fiber penetration of the package walls. The principles used to accomplish this concept involves a second-order grating to couple light in or out of the photonic circuit, and a binary optic lens which receives this light and focuses it into a single-mode optical fiber. Design, fabrication and electrical/optical measurements of this packaging concept are described.

An Advanced Fork System has been designed to add gamma-ray collimation and spectroscopy capability to the Fork measurement system, which has been used for burnup verification at pressurized water reactors (PWR). The Advanced Fork System measures the neutron and gamma-ray yields and the energy spectnum of gamma-rays from spent fuel assemblies. A cadmium-zinc-telluride (CZT) crystal permits the identification of the radioactive isotopes of cesium (134 and 137). The cesium isotope concentrations, with proper calibration can be used to determine the assembly burnup independent of reactor records, and to provide a measure of minimum cooling time. Tungsten gamma-ray collimators are used to define the spatial resolution of the gamma-ray detectors along the axis of the assembly. The capability to rapidly perform a burnup distribution scan using the collimated ion chamber may be important to the verification of burnup for boiling water reactors (BWR).

An open-loop control method is presented for reducing the oscillatory motion of rotary crane payloads during operator commanded maneuvers. A typical rotary crane consists of a multiple degree-of-freedom platform for positioning a spherical pendulum with an attached payload. The crane operator positions the Payload by issuing a combination of translational and rotational commands to the platform as well as load-line length changes. Frequently, these pendulum modes are time-varying and exhibit low natural frequencies. Maneuvers are therefore performed at rates sufficiently slow so as not to excite oscillation. The strategy presented here generates crane commands which suppress vibration of the payload without a priori knowledge of the desired maneuver. Results are presented for operator in-the-loop positioning using a real-time dynamics simulation of a three-axis rotary crane where the residual sway magnitude is reduced in excess of 4OdB.

A real-time method is presented to adoptively estimate three-dimensional unmodeled external torques acting on a spacecraft. This is accomplished by forcing the tracking error dynamics to follow the Lyapunov function underlying the feedback control law. For the case where the external torque is constant, the tracking error dynamics are shown to converge asypmtotically. The methodology applies not only to the control law used in this paper, but can also be applied to most Lyapunov derived feedback control laws. The adaptive external torque estimation is very robust in the presence of measurement noise, since a numerical integration is used instead of a numerical differentiation. Spacecraft modeling errors, such as in the inertia matrix, are also compensated for by this method. Several examples illustrate the practical significance of these ideas.

It is often desirable to separate particles from a particle-laden fluid stream. This is typically accomplished by passing the stream through a filter, an impactor, or a cyclone. In each of these devices, particles encounter obstacles in the flow path (i.e. filter material, the impaction surface, the cyclone side wall). However, in some applications, it is desirable to prevent particles from impinging on solid surfaces. For example, particle interaction with a solid surface may contaminate the surface, modify the particles via mechanical or chemical processes, or adversely affect the surface via material modification or heat transfer. In such situations, it is still possible to separate particles from the particle-laden flow stream by transferring them to another adjacent flow stream. This transfer of particles from one flow stream to another is termed nonimpact particle separation. One type of device that separates particles from a flow stream by nonimpact particle separation is the anticyclone. In contradistinction to a cyclone, the particle-laden flow is deflected from its original direction by a wall that curves away from the original flow direction, rather than into it. The computational fluid dynamics code FIDAP (Fluid Dynamics International) is used to perform two-dimensional fluid-flow and particle-motion calculations for a representative device geometry. These calculations indicate that the anticyclone geometry examined accomplishes nonimpact particle separation, as expected. Flow patterns and overall particle-separation characteristics are found to be fairly insensitive to Reynolds number for values above 100 regardless of whether the flow is laminar or turbulent. An approximate analytical relation describing anticyclone nonimpact particle separation is developed and validated by comparison to the numerical simulations. The additional information required to design useful devices employing nonimpact particle separation is outlined.

This report has demonstrated techniques that can be used to construct solutions to the 3-D electromagnetic inverse problem using full wave equation modeling. To this point great progress has been made in developing an inverse solution using the method of conjugate gradients which employs a 3-D finite difference solver to construct model sensitivities and predicted data. The forward modeling code has been developed to incorporate absorbing boundary conditions for high frequency solutions (radar), as well as complex electrical properties, including electrical conductivity, dielectric permittivity and magnetic permeability. In addition both forward and inverse codes have been ported to a massively parallel computer architecture which allows for more realistic solutions that can be achieved with serial machines. While the inversion code has been demonstrated on field data collected at the Richmond field site, techniques for appraising the quality of the reconstructions still need to be developed. Here it is suggested that rather than employing direct matrix inversion to construct the model covariance matrix which would be impossible because of the size of the problem, one can linearize about the 3-D model achieved in the inverse and use Monte-Carlo simulations to construct it. Using these appraisal and construction tools, it is now necessary to demonstrate 3-D inversion for a variety of EM data sets that span the frequency range from induction sounding to radar: below 100 kHz to 100 MHz. Appraised 3-D images of the earth`s electrical properties can provide researchers opportunities to infer the flow paths, flow rates and perhaps the chemistry of fluids in geologic mediums. It also offers a means to study the frequency dependence behavior of the properties in situ. This is of significant relevance to the Department of Energy, paramount to characterizing and monitoring of environmental waste sites and oil and gas exploration.

In the stereolithography process, three dimensional parts are built layer by layer using a laser to selectively cure slices of a photocurable resin, one on top of another. As the laser spot passes over the surface of the resin, the ensuing chemical reaction causes the resin to shrink and stiffen during solidification. When laser paths cross or when new layers are cured on top of existing layers, residual stresses are generated as the cure shrinkage of the freshly gelled resin is constrained by the adjoining previously-cured material. These internal stresses can cause curling in the compliant material. A capability for performing finite element analyses of the stereolithography process has been developed. Although no attempt has been made to incorporate all the physics of the process, a numerical platform suitable for such development has been established. A methodology and code architecture have been structured to allow finite elements to be birthed (activated) according to a prescribed order mimicking the procedure by which a laser is used to cure and build-up surface layers of resin to construct a three dimensional geometry. In its present form, the finite element code incorporates a simple phenomenological viscoelastic material model of solidification that is based on the shrinkage and relaxation observed following isolated, uncoupled laser exposures. The phenomenological material model has been used to analyze the curl in a simple cantilever beam and to make qualitative distinctions between two contrived build styles.

Sandia National Laboratories, New Mexico, conducts the Utility Battery Storage Systems Program, which is sponsored by the U.S. Department of Energy`s Office of Utility Technologies. The goal of this program is to assist industry in developing cost-effective battery systems as a utility resource option by 2000. Sandia is responsible for the engineering analyses, contracted development, and testing of rechargeable batteries and systems for utility energy storage applications. This report details the technical achievements realized during fiscal year 1995.

This report briefly describes the improvements and corrections made to a seismic network performance modeling code called NetSim. After expanding its supporting database, the modified code was used to predict the detection and location performances of an International Seismic Monitoring System (ISMS) proposed early in 1995 by the Group of Scientific Experts (GSE) supporting the Comprehensive Test Ban Treaty (CTBT) negotiators. The performance estimates indicated that the primary network achieves or exceeds the GSE goal of detecting a fully coupled 1kiloton nuclear shot contained in consolidated rock anywhere on the earth. It was also shown that without calibrated regional location models, the primary network alone cannot achieve the GSE location accuracy requirement of 1000 square kilometers at the detection threshold of the primary network, but with the help of 67 auxiliary stations that goal can be achieved in the northern hemisphere. Once the regional location models become better calibrated, the predictions based on these models indicated that the primary network should be able to realize the GSE requirement throughout much of the world. However, the location accuracy requirement is not realized in Central America, on the oceanic islands, on continental margins in the southern hemisphere, and in Antarctica. The introduction of the 67 auxiliary stations into the calibrated network primarily broadens the regions already having good location accuracies. The location performance estimates may be regarded as conservative since the S-wave arrivals were not included A comprehensive set of scenarios are illustrated in this effort to better understand various influences on network performance.

The end of the cold war has resulted in many changes for the Nuclear Weapons Complex (NWC). We now work in a smaller complex, with reduced resources, a smaller stockpile, and no new phase 3 weapons development programs. This new environment demands that we re-evaluate the way we design and produce nuclear weapons. The Defense Program (DP) Business Practices Re-engineering activity was initiated to improve the design and production efficiency of the DP Sector. The activity had six goals: (1) to identify DP business practices that are exercised by the Product Realization Process (PRP); (2) to determine the impact (positive, negative, or none) of these practices on defined, prioritized customer criteria; (3) to identify business practices that are candidates for elimination or re-engineering; (4) to select two or three business practices for re-engineering; (5) to re-engineer the selected business practices; and (6) to exercise the re-engineered practices on three pilot development projects. Business practices include technical and well as administrative procedures that are exercised by the PRP. A QFD exercise was performed to address (1)-(4). The customer that identified, defined, and prioritized the criteria to rate the business practices was the Block Change Advisory Group. Five criteria were identified: cycle time, flexibility, cost, product performance/quality, and best practices. Forty-nine business practices were identified and rated per the criteria. From this analysis, the group made preliminary recommendations as to which practices would be addressed in the re-engineering activity. Sixteen practices will be addressed in the re-engineering activity. These practices will then be piloted on three projects: (1) the Electronic Component Assembly (ECA)/Radar Project, (2) the B61 Mod 11, and (3) Warhead Protection Program (WPP).

We report measurements of the energies and angular distributions of positive ions in an inductively coupled argon plasma in a GEC reference cell. Use of two separate ion detectors allowed measurement of ion energies and fluxes as a function of position as well as ion angular distributions on the discharge centerline. The inductive drive on our system produced high plasma densities (up to 10{sup 12}/cm{sup 3} electron densities) and relatively stable plasma potentials. As a result, ion energy distributions typically consisted of a single feature well separated from zero energy. Mean ion energy was independent of rf power and varied inversely with pressure, decreasing from 29 eV to 12 eV as pressure increased form 2.4 m Torr to 50 mTorr. Half-widths of the ion angular distributions in these experiments varied from 5 degrees to 12.5 degrees, or equivalently, transverse temperatures varied form 0.2 to 0.5 eV with the distributions broadening as either pressure or RF power were increased.

Slide coating flow is a workhorse process for manufacturing precision film-coating products. Properly starting up a slide coating process is very important in reducing wastage during startup and ensuring that the process operates within the desired `coating window.` A two-phase flow analysis of slide-coating startup was performed by Palmquist and Scriven (1994) using Galerkin`s method with finite-element basis functions and an elliptic mesh generation scheme. As reported by Chen (1992) from flow visualization experiments, a continuously coated liquid film breaks up into rivulets, which are coating stripes with dry lanes in between, when the coated film becomes thinner and thinner due to either the increase in substrate speed or the reduction in pre-metered feed-liquid pump speed. It was observed that the coated-film breakup process originated from the coating bead, thus the name of bead breakup. Understanding the bead-breakup phenomena and elucidating mechanisms involved will provide guidance for manufacturing thinner coating, an industrial trend for better product performance. In this paper we present simulation results of slide-coating flows obtained from a computational method capable of describing arbitrary, three-dimensional and time-dependent deformations. The method, which is available in a commercial code, uses a fixed grid through which fluid interfaces are tracked by a Volume-of-Fluid technique (Hirt and Nichols, 1981). Surface tension, wall adhesion, and viscous stresses are fully accounted for in our analysis. We illustrate our computational approach by application to startup and the bead-breakup problems. As will be shown, for rapid processes our approach offers the computational efficiency and robustness that are difficult o achieve in conventional finite-element-based methods.

We detail a new monolithically integrated silicon mold/surface-micromachining process which makes possible the fabrication of stiff, high-aspect-ratio micromachined structures integrated with finely detailed, compliant structures. An important example, which we use here as our process demonstration vehicle, is that of an accelerometer with a large proof mass and compliant suspension. The proof mass is formed by etching a mold into the silicon substrate, lining the mold with oxide, filling it with mechanical polysilicon, and then planarizing back to the level of the substrate. The resulting molded structure is recessed into the substrate, forming a planar surface ideal for subsequent processing. We then add surface-micromachined springs and sense contacts. The principal advantage of this new monolithically integrated mold/surface-micromachining process is that it decouples the design of the different sections of the device: In the case of a sensitive accelerometer, it allows us to optimize independently the proof mass, which needs to be as large, stiff, and heavy as possible, and the suspension, which needs to be as delicate and compliant as possible. The fact that the high-aspect-ratio section of the device is embedded in the substrate enables the monolithic integration of high-aspect-ratio parts with surface-micromachined mechanical parts, and, in the future, also electronics. We anticipate that such an integrated mold/surface micromachining/electronics process will offer versatile high-aspect-ratio micromachined structures that can be batch-fabricated and monolithically integrated into complex microelectromechanical systems.