Publications

High-density plasma etching has been an effective patterning technique for the group-III nitrides due to ion fluxes which are 2 to 4 orders of magnitude higher than more conventional reactive ion etch (RIE) systems. GaN etch rates exceeding 0.68 μm/min have been reported in Cl2/H2/Ar inductively coupled plasmas (ICP) at -280 V dc-bias. Under these conditions, the etch mechanism is dominated by ion bombardment energies which can induce damage and minimize etch selectivity. High selectivity etch processes are often necessary for heterostructure devices which are becoming more prominent as growth techniques improve. In this study, we will report high-density ICP etch rates and selectivities for GaN, AlN, and InN as a function of cathode power, ICP-source power, and chamber pressure. GaN:AlN selectivities >8:1 were observed in a Cl2/Ar plasma at 10 m Torr pressure, 500 W ICP-source power, and 130 W cathode rf-power, while the GaN:InN selectivity was optimized at approximately 6.5:1 at 5 m Torr, 500 W ICP-source power, and 130 W cathode rf-power.

The mechanical properties of implanted layers and thin films on dissimilar substrates are difficult to determine accurately. Nanoindentation of the layer provides information, but detailed numerical modeling is required in order to separate the properties of the layer from those of the substrate. We describe here the procedures we have developed to accomplish this modeling with the commercially available finite-element code ABAQUS. Using these techniques, we are able to extract from nanoindentation testing the yield stress, Young's modulus, and hardness of the layer material, with an absolute accuracy of at least 20%. The procedure is applicable to layers as thin as 50 nm on essentially any substrate, hard or soft. We have used it for materials ranging from ion-implanted layers to thin films of metals and dielectrics formed using plasma-deposition methods. An example is given of O-implanted Al, a thin, hard layer on a soft substrate.

We have developed, prototyped, and demonstrated the feasibility of a novel robotic technique for rapid fabrication of composite structures. Its chief innovation is that, unlike all other available fabrication methods, it does not require a mold. Instead, the structure is built patch by patch, using a rapidly reconfigurable forming surface, and a robot to position the evolving part. Both of these components are programmable, so only the control software needs to be changed to produce a new shape. Hence it should be possible to automatically program the system to produce a shape directly from an electronic model of it. It is therefore likely that the method will enable faster and less expensive fabrication of composites.

Sandia National Laboratories is initiating development of a complete numerical simulation of parachute performance, beginning with parachute deployment and continuing through inflation and steady state descent. The purpose of the parachute performance code is to predict the performance of stockpile weapon parachutes as these parachutes continue to age well beyond their intended service life. A new massively parallel computer will provide unprecedented speed and memory for solving this complex problem, and new software will be written to treat the coupled fluid, structure and trajectory calculations as part of a single code. Verification and validation experiments have been proposed to provide the necessary confidence in the computations.

Fabrication of group-III nitride electronic and photonic devices relies heavily on the ability to pattern features with anisotropic profiles, smooth surface morphologies, etch rates often exceeding 0.5 μm/min, and a low degree of plasma-induced damage. Patterning these materials has been especially difficult due to their high bond energies and their relatively inert chemical nature as compared to other compound semiconductors. However, high-density plasma etching has been an effective patterning technique due to ion fluxes which are 2 to 4 orders of magnitude higher than conventional RIE systems. GaN etch rates as high as ≈1.3 μm/min have been reported in ECR generated ICl plasmas at-150V de-bias. In this study, we report high-density GaN etch results for ECR- and ICP-generated plasmas as a function of Cl2- and BCl3-based plasma chemistries.

Spurious behavior in underresolved grids and/or semiimplicit temporal discretizations for four computational fluid dynamics (CFD) simulations are studied. The numerical simulations consist of (a) a 1-D chemically relaxed nonequilibrium flow model, (b) the direct numerical simulation (DNS) of 2-D incompressible flow over a backward facing step, (c) a loosely-coupled approach for a 2-D fluid-structure interaction, and (d) a 3-D compressible unsteady flow simulation of vortex breakdown on delta wings. These examples were chosen based on their non-apparent spurious behaviors that were difficult to detect without extensive grid and/or temporal refinement studies and without some knowledge from dynamical systems theory. Studies revealed the various possible dangers of misinterpreting numerical simulation of realistic complex flows that are constrained by available computing power. In large scale computations underresolved grids, semi-implicit procedures, loosely-coupled implicit procedures, and insufficiently long time integration in DNS are most often unavoidable. Consequently, care must be taken in both computation and in interpretation of the numerical data. The results presented confirm the important role that dynamical systems theory can play in the understanding of the nonlinear behavior of numerical algorithms and in aiding the identification of the sources of numerical uncertainties in CFD.

Measurements of the threshold for secondary electron emission and shifts of the carbon Auger line position have been used to deduce the surface potential of several common phosphors during irradiation by electrons in the 0.5-5.0 keV range. All of the insulating phosphors display similar behavior: the surface potential is within ±1 V of zero at low electron energies. However, above 2-3 kV it becomes increasingly negative, reaching hundreds of volts within 1 keV of the turn-on energy. The electron energy at which this charging begins decreases dramatically after Coulomb aging at 17 μA/cm2 for 30-60 min. Measurements using coincident electron beams at low and high electron energies to control the surface potential were made to investigate the dependence of the cathodoluminescence (CL) process on charging. Initially, the CL from the two beams is identical to the sum of the separate beam responses, but after Coulomb aging large deviations from this additivity are observed. These results indicate that charging has important, detrimental effects on CL efficiency after prolonged e-beam irradiation. Measurements of the electron energy dependence of the CL efficiency before and after Coulomb aging will also be presented, and the implications of these data on the physics of the low-voltage CL process will be discussed.

A one-dimensional model is presented which describes the release-etch behavior of sacrificial oxides in aqueous HF. Starting from first principles and an empirical rate law, release etch kinetics are derived for primitive geometries. The behavior of complex three-dimensional structures is described by joining the solutions of constituent primitives and applying appropriate boundary conditions. The two fitting parameters, k1 and k2, are determined from the simplest structure and describe the more complex structures well. Experimental validation of the model is presented with data for all of the geometries and four types of sacrificial oxides.

Technology planning is becoming critical with the rapid development and obsolescence of technologies. Technology roadmapping provides a tool for selecting which technologies to pursue in what timeframes. This paper provides a framework for technology roadmaps, describes the roadmapping process, and reviews its application.

A method is described for generating electron cross sections that are compatible with standard discrete ordinates codes without modification. There are many advantages to using an established discrete ordinates solver, e.g., immediately available adjoint capability. Coupled electron-photon transport capability is needed for many applications, including the modeling of the response of electronics components to space and synthetic radiation environments. The cross sections have been successfully used in the DORT, TWODANT, and TORT discrete ordinates codes. The cross sections are shown to provide accurate and efficient solutions to certain multidimensional electron-photon transport problems. The key to the method is a simultaneous solution of the continuous-slowing-down and elastic-scattering portions of the scattering source by the Goudsmit-Saunderson theory. The resulting multigroup-Legendre cross sections are much smaller than the true scattering cross sections that they represent. Under certain conditions, the cross sections are guaranteed positive and converge with a low-order Legendre expansion.

This article attempts to review the progress achieved in the understanding of scaling and size effect in the failure of structures. Particular emphasis is placed on quasibrittle materials for which the size effect is complicated. Attention is focused on three main types of size effects, namely the statistical size effect due to randomness of strength, the energy release size effect, and the possible size effect due to fractality of fracture or microcracks. Definitive conclusions on the applicability of these theories are drawn. Subsequently, the article discusses the application of the known size effect law for the measurement of material fracture properties, and the modeling of the size effect by the cohesive crack model, nonlocal finite element models and discrete element models. Extensions to compression failure and to the rate-dependent material behavior are also outlined. The damage constitutive law needed for describing a microcracked material in the fracture process zone is discussed. Various applications to quasibrittle materials, including concrete, sea ice, fiber composites, rocks and ceramics are presented.

Two methods were examined for the fabrication of dielectric mirror masks. In the first method, a commercial laser mirror was patterned with photoresist and the dielectric film etched with ammonium bifluoride. The ammonium bifluoride etch showed strong kinetic anisotropy with the fastest etch rate in the vertical direction. However, horizontal etching still resulted in significant undercutting of the photomask. In the second method, a photoresist coated laser mirror was etched with an argon plasma. The argon plasma caused significant damage to the photoresist and underlying dielectric layer without adequate removal of the dielectric film in the open areas of the mask. Neither of the two methods examined were able to produce usable dielectric masks. During the course of this project, it was discovered that a foreign company, Balzers AG of Liechtenstein, had recently developed successful fabrication procedures for dielectric mirror masks. A mask purchased from Balzers for testing showed distinguishable pattern features down to 2 {mu}m in size. This mask was used in ablative projection etching experiments to form microstructures in Mylar polymer films. A thin film resistor pattern with 7.0 {mu}m wide lines was etched 5.4 {mu}m deep into a Mylar substrate. The etch pattern showed uniform linewidths but exhibited some thinning of the lines in areas where U-turns occurred. The ablative projection etching technique shows promise as a method for the rapid fabrication of contact masks in microstructuring applications.

A fatigue test of a wind turbine blade was conducted at the National Renewable Energy Laboratory in the fall of 1994. Acoustic emission monitoring of the test was performed, starting with the second loading level. The acoustic emission data indicated that this load exceeded the strength of the blade. From the first cycle at the new load, an oil can type of deformation occurred in two areas of the upper skin of the blade. One of these was near the blade root and the other was about the middle of the tested portion of the blade. The emission monitoring indicated that no damage was taking place in the area near the root, but in the deforming area near the middle of the blade, damage occurred from the first cycles at the higher load. The test was stopped after approximately one day and the blade was declared destroyed, although no gross damage had occurred. Several weeks later the test was resumed, to be continued until gross damage occurred. The upper skin tore approximately one half hour after the cycling was restarted.

Quartz crystal microbalances (QCMs) are piezoelectric thickness-shear-mode resonators where the resonant frequency has long been known to vary linearly with the mass of rigid layers on the surface when the device is in contact with air. This reports summarizes the results from a Laboratory Directed Research and Development effort to use an array of QCMs to measure and identify volatile organic compounds (VOCs) in water solutions. A total of nine polymer-coated QCMs were tested with varying concentrations of twelve VOCs while frequency and damping voltage were measured. Results from these experiments were analyzed using a Sandia-developed pattern recognition technique called visually empirical region of influence (VERI) developed at Sandia. The VERI analyses of data with up to 16% and 50% sensitivity drifts were carried out on an array with six signals obtained from five sensors. The results indicate that better than 98% and 88% correct chemical recognition is maintained for the 16% and 50% drifts, respectively. These results indicate a good degree of robustness for these sensor films.

Understanding the mechanisms that impact the performance of Microelectromechanical Systems (MEMS) is essential to the development of optimized designs and fabrication processes, as well as the qualification of devices for commercial applications. Silicon micromachines include engines that consist of orthogonally oriented linear comb drive actuators mechanically connected to a rotating gear. These gears are as small as 50 {mu}m in diameter and can be driven at rotation rates exceeding 300,000 rpm. Optical techniques offer the potential for measuring long term statistical performance data and transient responses needed to optimize designs and manufacturing techniques. We describe the development of Micromachine Optical Probe (MOP) technology for the evaluation of micromachine performance. The MOP approach is based on the detection of optical signals scattered by the gear teeth or other physical structures. We present experimental results obtained with a prototype optical probe and micromachines developed at Sandia National Laboratories.

It is common practice in system analysis to develop mathematical models for system behavior. Frequently, the actual system being modeled is also available for testing and observation, and sometimes the test data are used to help identify the parameters of the mathematical model. However, no general-purpose technique exists for formally, statistically judging the quality of a model. This paper suggests a formal statistical procedure for the validation of mathematical models of systems when data taken during operation of the system are available. The statistical validation procedure is based on the bootstrap, and it seeks to build a framework where a statistical test of hypothesis can be run to determine whether or not a mathematical model is an acceptable model of a system with regard to user-specified measures of system behavior. The approach to model validation developed in this study uses experimental data to estimate the marginal and joint confidence intervals of statistics of interest of the system. These same measures of behavior are estimated for the mathematical model. The statistics of interest from the mathematical model are located relative to the confidence intervals for the statistics obtained from the experimental data. These relative locations are used to judge the accuracy of the mathematical model. An extension of the technique is also suggested, wherein randomness may be included in the mathematical model through the introduction of random variable and random process terms. These terms cause random system behavior that can be compared to the randomness in the bootstrap evaluation of experimental system behavior. In this framework, the stochastic mathematical model can be evaluated. A numerical example is presented to demonstrate the application of the technique.

A method for more efficiently utilizing the frequency bandwidth allocated for data transmission is presented. Current space and range communication systems use modulation and coding schemes that transmit 0.5 to 1.0 bits per second per Hertz of radio frequency bandwidth. The goal in this LDRD project is to increase the bandwidth utilization by employing advanced digital communications techniques. This is done with little or no increase in the transmit power which is usually very limited on airborne systems. Teaming with New Mexico State University, an implementation of trellis coded modulation (TCM), a coding and modulation scheme pioneered by Ungerboeck, was developed for this application and simulated on a computer. TCM provides a means for reliably transmitting data while simultaneously increasing bandwidth efficiency. The penalty is increased receiver complexity. In particular, the trellis decoder requires high-speed, application-specific digital signal processing (DSP) chips. A system solution based on the QualComm Viterbi decoder and the Graychip DSP receiver chips is presented.

Electrokinetic remediation of uranium-contaminated soil was studied in a series of laboratory-scale experiments in test cells with identical geometry using quartz sand at approximately 10 percent moisture content. Uranium, when present in the soil system as an anionic complex, could be migrated through unsaturated soil using electrokinetics. The distance that the uranium migrated in the test cell was dependent upon the initial molar ratio of citrate to uranium used. Over 50 percent of the uranium was recovered from the test cells using the citrate and carbonate complexing agents over of period of 15 days. Soil analyses showed that the uranium remaining in the test cells had been mobilized and ultimately would have been extracted. Uranium extraction exceeded 90 percent in an experiment that was operated for 37 days. Over 70 percent of the uranium was removed from a Hanford waste sample over a 55 day operating period. Citrate and carbonate ligand utilization ratios required for removing 50 percent of the uranium from the uranium-contaminated sand systems were approximately 230 moles ligand per mole uranium and 1320 moles ligand per mole uranium for the waste. Modifying the operating conditions to increasing the residence time of the complexants is expected to improved the utilization efficiency of the complexing agent.

Tera Computer and Sandia National Laboratories have completed a CRADA, which examined the Tera Multi-Threaded Architecture (MTA) for use with large codes of importance to industry and DOE. The MTA is an innovative architecture that uses parallelism to mask latency between memories and processors. The physical implementation is a parallel computer with high cross-section bandwidth and GaAs processors designed by Tera, which support many small computation threads and fast, lightweight context switches between them. When any thread blocks while waiting for memory accesses to complete, another thread immediately begins execution so that high CPU utilization is maintained. The Tera MTA parallel computer has a single, global address space, which is appealing when porting existing applications to a parallel computer. This ease of porting is further enabled by compiler technology that helps break computations into parallel threads. DOE and Sandia National Laboratories were interested in working with Tera to further develop this computing concept. While Tera Computer would continue the hardware development and compiler research, Sandia National Laboratories would work with Tera to ensure that their compilers worked well with important Sandia codes, most particularly CTH, a shock physics code used for weapon safety computations. In addition to that important code, Sandia National Laboratories would complete research on a robotic path planning code, SANDROS, which is important in manufacturing applications, and would evaluate the MTA performance on this code. Finally, Sandia would work directly with Tera to develop 3D visualization codes, which would be appropriate for use with the MTA. Each of these tasks has been completed to the extent possible, given that Tera has just completed the MTA hardware. All of the CRADA work had to be done on simulators.

Multichip modules (MCMs) containing power components need a substrate with excellent heat spreading capability to both avoid hot spots and to move dissipation heat toward the system heat sinks. Polycrystalline diamond is an excellent MCM heat spreading substrate but remains several orders of magnitude too expensive and somewhat more difficult to process than conventional mother-board materials. Today`s power MCMs concentrate on moderately priced silicon wafers and aluminum nitride ceramic with their improved thermal conductivity and good thermal expansion match to power semiconductor components in comparison to traditional alumina and printed wiring board materials. However, even silicon and AlN substrates are thermally challenged by designers needs. The authors report on the integral fabrication of micro-heat pipes embedded in silicon MCM substrates (5 x 5 cm) by the use of micromachined capillary wick structures and hermetic micro-cavities. This passive microstructure results in more than a 5 times improvement in heat spreading capability of the silicon MCM substrate over a large range of power densities and operating temperatures. Thus diamond-like cooling is possible at silicon prices.

Detecting object boundaries in the presence of cast shadows is a difficult task for machine vision systems. A new edge detector is presented which responds to shadow penumbras and abrupt object edges with distinguishable signals. The detector requires the use of spatially extended light sources and sufficient video resolution to resolve the shadow penumbras of interest. Detection of high frequency noise is suppressed without requiring image-dependent adjustment of signal thresholds. The ability of the edge operator to distinguish shadow penumbras from abrupt object boundaries while suppressing responses to high frequency noise and texture is illustrated with idealized shadow and object edge intensity profiles. Selective detection of object boundaries in a video scene with a cast shadow has also been demonstrated with this operator.

Modular fixturing kits are precisely machined sets of components used for flexible, short-turnaround construction of fixtures for a variety of manufacturing purposes. A modular vise is a parallel-jaw vise, where each jaw is a modular fixture plate with a regular grid of precisely positioned holes. A modular vise can be used to locate and hold parts for machining, assembly, and inspection tasks. To fixture a part, one places pins in some of the holes so that when the vise is closed, the part is reliably located and completely constrained. The modular vise concept can be adapted easily to the design of modular parallel-jaw grippers for robots. By attaching a grid plate to each jaw of a parallel-jaw gripper, the authors gain the ability to easily construct high-quality grasps for a wide variety of parts from a standard set of hardware. Wallack and Canny developed a previous algorithm for planning planar grasp configurations for the modular vise. In this paper, the authors expand this work to produce a 3-d fixture/gripper design tool. They describe several analyses added to the planar algorithm to improve its utility, including a three-dimensional grasp quality metric based on geometric and force information, three-dimensional geometric loading analysis, and inter-gripper interference analysis to determine the compatibility of multiple grasps for handing the part from one gripper to another. Finally, the authors describe two applications which combine the utility of modular vise-style grasping with inter-gripper interference: The first is the design of a flexible part-handling subsystem for a part cleaning workcell under development at Sandia National Laboratories; the second is the automatic design of grippers that support the assembly of multiple products on a single assembly line.

The goal of the Modular Weapon Control Unit (MWCU) program was to design and develop a reconfigurable weapon controller (programmer/sequencer) that can be adapted to different weapon systems based on the particular requirements for that system. Programmers from previous systems are conceptually the same and perform similar tasks. Because of this commonality and the amount of re-engineering necessary with the advent of every new design, the idea of a modular, adaptable system has emerged. Also, the controller can be used in more than one application for a specific weapon system. Functionality has been divided into a Processor Module (PM) and an Input/Output Module (IOM). The PM will handle all operations that require calculations, memory, and timing. The IOM will handle interfaces to the rest of the system, input level shifting, output drive capability, and detection of interrupt conditions. Configuration flexibility is achieved in two ways. First, the operation of the PM is determined by a surface mount Read-Only Memory (ROM). Other surface-mount components can be added or neglected as necessary for functionality. Second, IOMs consist of configurable input buffers, configurable output drivers, and configurable interrupt generation. Further, these modules can be added singly or in groups to a Processor Module to achieve the required I/O configuration. The culmination of this LDRD was the building of both Processor Module and Input/Output Module. The MWCU was chosen as a test system to evaluate Low-Temperature Co-fired Ceramic (LTCC) technology, desirable for high component density and good thermal characteristics.

A series of three 18.9 m diameter JP-4 pool fire experiments with a large (2.1 m X 4.6 m), flat plate calorimeter adjacent to the fuel pool were recently performed. The objectives of these experiments were to: (1) gain a better understanding of fire phenomenology, (2) provide empirical input parameter estimates for simplified, deterministic Risk Assessment Compatible Fire Models (RACFMs), (3) assist in continuing fire field model code validation and development, and (4) enhance the data base of fire temperature and heat flux to object distributions. Due to different wind conditions during each experiment, data were obtained for conditions where the plate was not engulfed, fully-engulfed and partially engulfed by the continuous flame zone. Results include the heat flux distribution to the plate and flame thermocouple temperatures in the vicinity of the plate and at two cross sections within the lower region of the continuous flame zone. The results emphasize the importance of radiative coupling (i.e. the cooling of the flames by a thermally massive object) and convective coupling (including object-induced turbulence and object/wind/flame interactions) in determining the heat flux from a fire to an object. The formation of a secondary flame zone on an object adjacent to a fire via convective coupling (which increases the heat flux by a factor of two) is shown to be possible when the object is located within a distance equal to the object width from the fire.

This study involved the evaluation and documentation of cases in which petroleum wellbores were enlarged beyond the nominal hole diameter as a consequence of erosion during exploratory drilling, particularly as a function of gas flow into the wellbore during blowout conditions. A primary objective was to identify analogs to potential wellbore enlargement at the Waste Isolation Pilot Plant (WIPP) during inadvertent human intrusion. Secondary objectives were to identify drilling scenarios associated with enlargement, determine the physical extent of enlargement, and establish the physical properties of the formation in which the enlargement occurred. No analogs of sufficient quality to establish quantitative limits on wellbore enlargement at the WIPP disposal system were identified. However, some information was obtained regarding the frequency of petroleum well blowouts and the likelihood that such blowouts would bridge downhole, self-limiting the surface release of disposal-system material. Further work would be necessary, however, to determine the conditions under which bridging could occur and the extent to which the bridging might be applicable to WIPP. In addition, data on casing sizes of petroleum boreholes in the WIPP vicinity support the use of a 12-{1/4} inch borehole size in WIPP performance assessment calculations. Finally, although data are limited, there was no evidence of significant wellbore enlargement in any of three blowouts that occur-red in wellbores in the Delaware Basin (South Culebra Bluff Unit No. 1, Energy Research and Development Administration (ERDA) 6, and WIPP 12).