Publications

Be is a ``marginal metal.`` The stable phase, hcp-Be, has a low Fermi-level density of states and very anisotropic structural and elastic properties, similar to a semiconductor`s. At the Be(0001) surface, surface states drastically increase the Fermi-level density of states. The different nature of bonding in bulk-Be and at the Be(0001) surface explains the large outward relaxation. The presence of surface states causes large surface core-level shifts by inducing a higher electrostatic potential in the surface layers and by improving the screening at the surface. The authors experimental and theoretical investigations of atomic vibrations at the Be(0001) surface demonstrate clearly that Be screening of atomic motion by the surface states makes the surface phonon dispersion fundamentally different from that of the bulk. Properties of Be(0001) are so different from those of the bulk that the surface can be considered a new ``phase`` of beryllium with unique electronic and structural characteristics. For comparison they also study Be(11{bar 2}0), a very open surface without important surface states. Be(11{bar 2}0) is the only clean s-p metal surface known to reconstruct (1 {times} 3 missing row reconstruction).

NMR and neutron diffraction measurements reveal that macroscopic phase separation and the tetragonal to orthorhombic (TO) structural phase coincide at two distinct points in the temperature-doping phase plot for oxygen doped La{sub 2}CuO{sub 4+{delta}}. Thus the TO phase line coincides with the phase separation line. This is evidence that the macroscopic phase separation is inhibited in the tetragonal phase. We propose that the interstitial oxygen has higher mobility in the orthorhombic phase and that insufficient mobility suppresses macroscopic phase separation in the tetragonal phase. Neutron diffraction measurements also reveal superlattice peaks which indicate ordering of the interstitial oxygen. Our NMR measurements, have demonstrated a distribution of tilts of the CuO{sub 6} octahedra. We propose a sawtooth modulation of the octahedral tilt in which the sign of the tilt changes when the tilt reaches a maximum value can explain this distribution. The large openings in the La-O layer resulting from the abrupt switch of the sign of the tilt provide an attractive location for the interstitial oxygen. This mechanism would lead to stripe ordering of the interstitial oxygen.

Lateral and vertical variabilities in the bulk and mechanical properties of silicic volcanic tuff at the potential nuclear waste repository site in Yucca Mountain, NV have been evaluated. Laboratory measurements have been performed on tuff specimens recovered from boreholes located to support the design of the Exploratory Studies Facility/North Ramp. The data include dry and saturated bulk densities, average grain density, porosity, compressional and shear wave velocities, elastic moduli, and compressional and tensional fracture strengths. Data from eight boreholes aligned in a northwest-southeast direction have been collected under the required quality assurance program. Three boreholes have penetrated the potential repository horizon. The information collected provides for an accurate appraisal of the variability of rock properties in the vicinity of the boreholes. As expected, there is substantial variability in the bulk and mechanical properties of the tuff with depth (lithology). This is due to variations in gross characteristics of the tuffs (e.g., cooling units, mode of deposition, etc.), as well as smaller scale features (welding, porosity, and internal structures) that have developed as a result of depositional and post-depositional mechanisms. An evaluation of the lateral variability in bulk and mechanical properties is somewhat limited, at this time, due to a lack of borehole control to the north and south (parallel to the depositional flow direction). Initial observations indicate that there is minimal lateral variability within lithologic units. There are observable differences however, that can be related to variability in specific properties (e.g., porosity, and internal structures).

Short-wavelength, visible-light emitting optoelectronic devices are needed for a wide range of commercial applications, including high-density optical data storage, full-color displays, and underwater communications. In 1994, high-brightness blue LEDs based on gallium nitride and related compounds (InGaN/AlGaN) were introduced by Nichia Chemical Industries. The Nichia diodes are 100 times brighter than the previously available SiC blue LEDs. Group-III nitrides combine a wide, direct bandgap with refractory properties and high physical strength. So far, no studies of degradation of GaN based LEDs have been reported. The authors study, reported in this paper, focuses on the performance of GaN LEDs under high electrical stress conditions. Their observations indicate that, in spite of a high defect density, which normally would have been fatal to other III-V devices, defects in group-III nitrides are not mobile even under high electrical stress. Defect tubes, however, can offer a preferential path for contact metals to electromigrate towards the p-n junction, eventually resulting in a short. The proposed mechanism of GaN diode degradation raises concern for prospects of reliable lasers in the group-III nitrides grown on sapphire.

There is uncertainty in the performance of wind energy installations due to unknowns in the local wind environment, machine response to the environment, and the durability of materials. Some of the unknowns are inherently independent from machine to machine while other uncertainties are common to the entire fleet equally. The FAROW computer software for fatigue and reliability of wind turbines is used to calculate the probability of component failure due to a combination of all sources of uncertainty. Although the total probability of component failure due to all effects is sometimes interpreted as the percentage of components likely to fail, this perception is often far from correct. Different amounts of common versus independent uncertainty are reflected in economic risk due to either high probabilities that a small percentage of the fleet will experience problems or low probabilities that the entire fleet will have problems. The average, or expected cost is the same as would be calculated by combining all sources of uncertainty, but the risk to the fleet may be quite different in nature. Present values of replacement costs are compared for two examples reflecting different stages in the design and development process. Results emphasize that an engineering effort to test and evaluate the design assumptions is necessary to advance a design from the high uncertainty of the conceptual stages to the lower uncertainty of a well engineered and tested machine.

The Photovoltaic Manufacturing Technology (PVMaT) Project was initiated in 1990 to help the US photovoltaic (PV) industry extend its world leadership role in manufacturing and commercially developing PV modules and systems. It is being conducted in several phases, staggered to support industry progress. The four most recently awarded subcontracts (Phase 2B) are now completing their first year of research. They include two subcontracts on CdTe, one on Spheral Solar[trademark] Cells, and one on cast polysilicon. These subcontracts represent new technology additions to the PVMaT Project. Subcontracts initiated in earlier phases are nearing completion, and their progress is summarized. An additional phase of PVMaT, Phase 4A, is being initiated which will emphasize product-driven manufacturing research and development. The intention of Phase 4A is to emphasize improvement and cost reduction in the manufacture of full-system PV products. The work areas may include, but are not limited to, issues such as improvement of module manufacturing processes; system and system component packaging, integration, manufacturing, and assembly; product manufacturing flexibility; and balance-of-system development with the goal of product manufacturing improvements.

In coating processes (e.g. in blade coating) the flow domain inherently contains free surfaces and three-phase contact lines, and characteristic length scales of flow features in the dimension transverse to the web-movement vary by an order of magnitude or more from a fraction of a millimeter or more to tens of microns or less). The presence of free surfaces and three-phase contact lines, and the sudden changes of flow geometry and directions create difficulties in theoretical analyses of such flows. Though simulations of coating flows via finite-element methods using structured grids have been reportedly demonstrated in the literature, achieving high efficiency of such numerical experiments remains a grand challenge -- mainly due to difficulties in local mesh-refinement and in avoiding unacceptably distorted grids. High efficiency of computing steady flow fields under various process conditions is crucial in shortening turn-around time in design and optimization of coating-flow processes. In this paper we employ a fully-implicit, pseudo-solid, domain mapping technique coupled with unstructured meshes to analyze blade and slot coating flows using Galerkin`s method with finite element basis functions. We demonstrate the robustness and efficiency of our unique technique in circumventing shortcomings of mesh-motion schemes currently being used in the coating-flow research community. Our goal is to develop an efficient numerical tool, together with a suitable optimization toolkit, that can be used routinely in design and optimization of coating-flow processes.

Radionuclide releases due to drilling into the potential Yucca Mountain nuclear-waste repository have been evaluated as part of a recent total-system performance assessment. The probability that a drilling event intersects a waste package is a function of the sizes of the drill bit and the waste package, and the density of placement of the containers in the repository. The magnitude of the releases is modeled as a random function that also depends on the amount of decay the radionuclides have undergone. Four cases have been analyzed, representing the combinations of two waste-package designs (small-capacity, thin-wall, vertically emplaced; and large-capacity, thick-wall, horizontally emplaced) and two repository layouts (lower thermal power dissipation, low waste-package placement density; and higher thermal power dissipation, high waste-package placement density). The results show a fairly pronounced dependence on waste-package design and slight dependence on repository layout. Given the assumptions in the model, releases from the larger containers are 4-5 times greater than from the smaller packages.

When a modal test is to be performed for purposes of correlation with a finite element model, one needs to design the test so that the resulting measurements will provide the data needed for the correlation. There are numerous issues to consider in the design of a modal test; two important ones are the number and location of response sensors, and the number, location, and orientation of input excitation. From a model correlation perspective, one would like to select the response locations to allow a definitive, one-to-one correspondence between the measured modes and the predicted modes. Further, the excitation must be designed to excite all the modes of interest at a sufficiently high level so that the modal estimation algorithms can accurately extract the modal parameters. In this paper these two issues are examined in the context of model correlation with methodologies presented for obtaining an experiment design.

A technique which involves spincoating in conjunction with a thermodynamic process was used to develop microporous polyimide films with lower dielectric constant and better stress reduction properties. A soluble polyimide solution was spincoated into a silicon wafer, and then immediately submerged into toluene. The factors affecting film morphology, thickness, pore size, and percent porosity were investigated.

Prosperity Games are an outgrowth and adaptation of move/countermove and seminar War Games. Prosperity Games are simulations that explore complex issues in a variety of areas including economics, politics, sociology, environment, education and research. These issues can be examined from a variety of perspectives ranging from a global, macroeconomic and geopolitical viewpoint down to the details of customer/supplier/market interactions in specific industries. All Prosperity Games are unique in that both the game format and the player contributions vary from game to game. This report documents a 90-minute Prosperity Game conducted as part of Advanced Manufacturing Day on May 17, 1994. This was the fourth game conducted under the direction of the Center for National Industrial Alliances at Sandia. Although previous games lasted from one to two days, this abbreviated game produced interesting and important results. Most of the strategies proposed in previous games were reiterated here. These included policy changes in international trade, tax laws, the legal system, and the educational system. Government support of new technologies was encouraged as well as government-industry partnerships. The importance of language in international trade was an original contribution of this game. The deliberations and recommendations of these teams provide valuable insights as to the views of this diverse group of decision makers concerning policy changes, foreign competition, and the development, delivery and commercialization of new technologies.

Feature extraction transforms an object's image representation to an alternate reduced representation. In one-class object recognition, we would like this alternate representation to give improved discrimination between the object and all possible non-objects and improved generalization between different object poses. Feature selection can be time-consuming and difficult to optimize so we have investigated unsupervised neural networks for feature discovery. We first discuss an inherent limitation in competitive type neural networks for discovering features in gray level images. We then show how Sanger's Generalized Hebbian Algorithm (GHA) removes this limitation and describe a novel GHA application for learning object features that discriminate the object from clutter. Using a specific example, we show how these features are better at distinguishing the target object from other non-target objects with Carpenter's ART 2-A as the pattern classifier.

Verification measurements may be used to help ensure nuclear criticality safety when burnup credit is applied to spent fuel transport and storage systems. The FORK measurement system, designed at Los Alamos National Laboratory for the International Atomic Energy Agency safeguards program, has been used to verify reactor site records for burnup and cooling time for many years. The FORK system measures the passive neutron and gamma-ray emission from spent fuel assemblies while in the storage pool. This report deals with the application of the FORK system to burnup credit operations based on measurements performed on spent fuel assemblies at the Oconee Nuclear Station of Duke Power Company.

Lead zirconate titanate (PZT) thin films are of technological interest for a variety of electronic and optical applications. Fabrication of PZT films by solution deposition techniques is attractive because of stoichiometric control at the molecular level, ease of processing, and both low capital investment and total cost. Control of phase evolution, microstructure, crystallite size and orientation, and ferroelectric domain assemblage during processing is essential to optimize electrical and/or optical properties of the films. Electron microscopy techniques have been used extensively to correlate microstructural features with film processing.

This paper addresses theoretical aspects of forming images from an airborne Synthetic Aperture Radar (SAR) of targets buried below the earth's surface. Soil is generally a lossy, dispersive medium, with wide ranging variability in these attributes depending on soil type, moisture content, and a host of other physical properties. Focussing a SAR subsurface image presents new dimensions of complexity relative to its surface-image counterpart, even when the soil's properties are known. This paper treats the soil as a lossy, dispersive half space, and presents a practical model for the radar echo-delay time to point scatterers within it. This model is then used to illustrate effects of refraction, dispersion, and attenuation on a SAR's phase histories, and the resulting image. Various data collection geometries and processing strategies are examined for both 2-Dimensional and 3-Dimensional SAR images. The conclusions from this work are that 1) focussing a SAR image must generally take into account both refraction and dispersion, 2) resolving targets at different depths in lossy soils requires perhaps unprecedented sidelobe attenuation, that for some soils may only be achievable with specialized window functions, 3) the impulse response of the soil itself places a practical limit on the usable bandwidth of the radar, and 4) dynamic ranges and sensitivities will need to be orders of magnitude greater than typical surface-imaging SARs, leading to significant impact on SAR parameters, for example compressing the usable range of pulse repetition frequencies (PRFs).

This paper demonstrates how certain concepts from the Phase Gradient Autofocus (PGA) algorithm for automated refocus of spotlight mode SAR imagery may be used to design a similar algorithm that applies to SAR imagery formed in the conventional strip-mapping mode. The algorithm derivation begins with the traditional view of strip-map image formation as convolution (compression) using a linear FM chirp sequence. The appropriate analogies and modifications to the spotlight mode case are used to describe a working algorithm for strip-map autofocus.

Statistical methods were used to design and analyze the results of a gettering experiment on four industrial multicrystalline silicon solar cell materials. The experiment studied the effects of temperature and time in the POCl3 diffusion process and the aluminum alloy process using simple diagnostic devices. The time and temperature ranges were restricted to maintain compatibility with commercial fabrication sequences. The design was capable of picking up second order interactions between the various processing factors. Statistically significant gettering effects were detected in only two of the four materials. The results for one of these materials were further tested using full solar cells. Strengths and weaknesses of this approach to gettering studies have become apparent in the present work and are discussed.

Computational physicists at Sandia National Laboratories have moved their production codes to distributed memory parallel computers. Such an effort required the development of parallel algorithms, parallel data bases and parallel support tools. The Eulerian CTH code was rewritten. Moving both ALEGRA and PRONTO to parallel computers required only a modest number of modifications. It involved restructuring the restart and graphics data bases to make them parallel and minimize the I/O to the parallel computer. It also involved developing mesh decomposition tools to divide a rectangular or arbitrary connectivity into sub-meshes. It also involved developing new visualization tools to process the very large, parallel data bases. This paper also discusses Sandia's experiences running these codes on its 1840 compute node Intel Paragon, 1024 processor nCUBE and networked stations.

Techniques for lossless waveform compression can be applied to the transmission of weight vectors from an orbiting satellite. The vectors, which are a part of a hybrid analog/digital adaptive filter, are a representation of the radio frequency background seen by the satellite. An approach is used which treats each adaptive weight as a time-varying waveform.

Transformers models for simulation with Pspice and Analogy's Saber are being developed using experimental B-H Loop and network analyzer measurements. The models are evaluated for accuracy and convergence using several test circuits. Results are presented which demonstrate the effects on circuit performance from magnetic core losses, eddy currents, and mechanical stress on the magnetic cores.

This paper describes the result of a team effort at Sandia to demonstrate the near-term performance potential for multicrystalline silicon modules using commercial mc-Si material and improved cell fabrication processes. Large-area high-performance mc-Si cells were fabricated, prototype modules were built, and world-record module efficiency was confirmed by outdoor testing at over 15% for standard test conditions.

A significant number of analytical problems (for example, abnormal-environment safety analysis) depend on data that are partly or mostly subjective. Since fuzzy algebra depends on subjective operands, we have been investigating its applicability to these forms of assessment, particularly for portraying uncertainty in the results of PRA (probabilistic risk analysis) and in risk-analysis-aided decision-making. Since analysis results can be a major contributor to a safety-measure decision process, risk management depends on relating uncertainty to only known (not assumed) information. The uncertainties due to abnormal environments are even more challenging than those in normal-environment safety assessments; and therefore require an even more judicious approach. Fuzzy algebra matches these requirements well. One of the most useful aspects of this work is that we have shown the potential for significant differences (especially in perceived margin relative to a decision threshold) between fuzzy assessment and probabilistic assessment based on subtle factors inherent in the choice of probability distribution models. We have also shown the relation of fuzzy algebra assessment to ``bounds`` analysis, as well as a description of how analyses can migrate from bounds analysis to fuzzy-algebra analysis, and to probabilistic analysis as information about the process to be analyzed is obtained. Instructive examples are used to illustrate the points.

Many organizations face high consequence safety situations where unwanted stimuli due to accidents, catastrophes, or inadvertent human actions can cause disasters. In order to improve interaction among such organizations and to build on each others` experience, preventive approaches, and assessment techniques, the High Consequence Operations Safety Symposium was held July 12--14, 1994 at Sandia National Laboratories, Albuquerque, New Mexico. The symposium was conceived by Dick Schwoebel, Director of the SNL Surety Assessment Center. Stan Spray, Manager of the SNL System Studies Department, planned strategy and made many of the decisions necessary to bring the concept to fruition on a short time scale. Angela Campos and about 60 people worked on the nearly limitless implementation and administrative details. The initial symposium (future symposia are planned) was structured around 21 plenary presentations in five methodology-oriented sessions, along with a welcome address, a keynote address, and a banquet address. Poster papers addressing the individual session themes were available before and after the plenary sessions and during breaks.

This report is a critical reassessment of the geotechnical risks of continuing oil storage at the Weeks Island Strategic Petroleum Reserve site. It reviews all previous risk abatement recommendations, subsequent mitigative actions, and new information. Of increased concern, due to the discovery of a surface levels, is the long term maintainability of the mine as an oil storage repository. Mine operational changes are supported in order to facilitate monitoring of water entry diagnostics. These changes are also intended to minimize the volume in the mine available for water entry. Specific recommendations are made to implement the mine changes.

The design of structural containments for testing small explosive devices requires the designer to consider the various aspects of the explosive loading, i.e., shock and gas or quasistatic pressure. Additionally, if the explosive charge has the potential of producing damaging fragments, provisions must be made to arrest the fragments. This may require that the explosive be packed in a fragment attenuating material, which also will affect the loads predicted for containment response. Material also may be added just to attenuate shock, in the absence of fragments. Three charge weights are used in the design. The actual charge is used to determine a design fragment. Blast loads are determined for a {open_quotes}design charge{close_quotes}, defined as 125% of the operational charge in the explosive device. No yielding is permitted at the design charge weight. Blast loads are also determined for an over-charge, defined as 200% of the operational charge in the explosive device. Yielding, but no failure, is permitted at this over-charge. This guide emphasizes the calculation of loads and fragments for which the containment must be designed. The designer has the option of using simplified or complex design-analysis methods. Examples in the guide use readily available single degree-of-freedom (sdof) methods, plus static methods for equivalent dynamic loads. These are the common methods for blast resistant design. Some discussion of more complex methods is included. Generally, the designer who chooses more complex methods must be fully knowledgeable in their use and limitations. Finally, newly fabricated containments initially must be proof tested to 125% of the operational load and then inspected at regular intervals. This specification provides guidance for design, proof testing, and inspection of small explosive containment structures.