Publications

Paramount to the modeling of unsaturated flow and transport through fractured porous media is a clear understanding of the processes controlling fracture-matrix interaction. As a first step toward such an understanding, two preliminary experiments have been performed to investigate the influence of matrix imbibition on water percolation through unsaturated fractures in the plane normal to the fracture. Test systems consisted of thin slabs of either tuff or an analog material cut by a single vertical fracture into which a constant fluid flux was introduced. Transient moisture content and solute concentration fields were imaged by means of x-ray absorption. Flow fields associated with the two different media were significantly different owing to differences in material properties relative to the imposed flux. Richards` equation was found to be a valid means of modeling the imbibition of water into the tuff matrix from a saturated fracture for the current experiment.

A set of detailed geostatistical simulations of porosity has been produced for a layered stratigraphic sequence of welded and nonwelded volcanic tuffs at Yucca Mountain, Nevada. The simulations are produced using a composite. model of spatial continuity and they are highly conditioned to abundant drill hole (core) information. A set of derivative simulations of saturated hydraulic conductivity has been produced, in the absence of conditioning data, using a cross-variable relationship developed from similar data elsewhere. The detailed simulations reproduce both the major stratigraphic units and finer scale layering indicated by the drill hole data. These simulations have been scaled up several order of magnitude to represent block-scale effective hydrologic properties suitable for use in numerical modeling of groundwater flow and transport. The upscaling process involves the reformulation of a previously reported method that iteratively adapts an initial arbitrary grid to ``homogenize`` the detailed hydraulic properties contained within the adjusted cell limits and to minimize the size of cell in highly heterogeneous regions. Although the computation of the block-effective property involves simple numerical averaging, the blocks over which these averages are computed are relatively homogeneous, which reduces the numerical difficulties involved in averaging non-additive properties, such as permeability. The entire process of simulation and upscaling is rapid and computationally efficient compared with alterative techniques. It is thus suitable for the Monte Carlo evaluation of the uncertainty in site characterization as it affects the results of groundwater flow and transport calculations.

Hydrologic properties have been measured on outcrop samples taken from a detailed, two-dimension grid covering a 1.4 km outcrop exposure of the 10-m thick non-welded-to-welded, shardy base microstratigraphic unit of the Tiva Canyon Member of the Miocene Paintbrush Tuff at Yucca Mountain, Nevada. These data allow quantification of spatial trends in rock matrix properties that exist in this important hydrologic unit. Geologic investigation, combined with statistical and geostatistical analyses of the numerical data, indicates that spatial variability of matrix properties is related to deterministic geologic processes that operated throughout the region. Linear vertical trends in hydrologic properties are strongly developed in the shardy base microstratigraphic unit, and they are more accurately modeled using the concept of a thickness-normalized stratigraphic elevation within the unit, rather than absolute elevation. Hydrologic properties appear to be correlated over distances of 0.25 to 0.3 of the unit thickness after removing the deterministic vertical trend. The use of stratigraphic elevation allows scaling of identified trends by unit thickness which may be of particular importance in a basal, topography-blanketing unit such as this one. Horizontal changes in hydrologic properties do not appear to form obvious trends within the limited lateral geographic extent of the ash-flow environment that was examined. Matrix properties appear to be correlated horizontally over distances between 100 and 400 m. The existence and quantitative description of these trends and patterns of vertical spatial continuity should increase confidence in models of hydrologic properties and groundwater flow in this area that may be constructed to support the design of a potential high-level nuclear waste repository at Yucca Mountain.

This paper presents a method of estimating the rock mass properties for the welded and nonwelded tuffs based on currently available information on intact rock and joint characteristics at the Yucca Mountain site. Variability of the expected ground conditions at the potential repository horizon (the TSw2 thermomechanical unit) and in the Calico Hills nonwelded tuffs is accommodated by defining five rock mass quality categories in each unit based upon assumed and observed distributions of the data.

A numerical approach for modeling unsaturated flow is developed for heterogeneous simulations of fractured tuff generated using a geostatistical method. Cross correlations of hydrologic properties and upscaling of moisture retention curves is discussed. The approach is demonstrated for a study of infiltration at Yucca Mountain.

The regulations that currently govern repositories for spent fuel and high-level waste require demonstrations that are sometimes described as impossible to make. To make them will require an understanding of the current and the future phenomena at repository sites; it will also require credible estimates of the probabilities that the phenomena will occur in the distant future. Experts in many fields{emdash}earth sciences, statistics, numerical modeling, and the law{emdash}have questioned whether any amount of data collection can allow modelers to meet these requirements with enough confidence to satisfy the regulators. In recent years some performance assessments have begun to shed light on this question because they use results of actual site investigations. Although these studies do not settle the question definitively, a review of a recent total-system assessment suggests that compliance may be possible to demonstrate. The review also suggests, however, that the demonstration can be only at the ``reasonable`` levels of assurance mentioned, but not defined, in the regulations.

Previous laboratory investigations of tuff have shown that porosity has a dominant, general effect on mechanical properties. As a result, it is very important for the interpretation of mechanical property data that porosity is measured on each sample tested. Porosity alone, however, does not address all of the issues important to mechanical behavior. Variability in size and distribution of pore space produces significantly different mechanical properties. A nondestructive technique for characterizing the internal structure of the sample prior to testing is being developed and the results are being analyzed. The information obtained from this technique can help in both qualitative and quantitative interpretation of test results.

In situ thermomechanical experiments are planned as part of the Yucca Mountain Site Characterization Project that require instruments to measure stress and displacement at temperatures that exceed the typical specifications of existing geotechnical instruments. A high degree of instrument reliability will also be required to satisfy the objectives of the experiments, therefore a study was undertaken to identify areas where improvement in instrument performance was required. A preliminary list of instruments required for the experiments was developed, based on existing test planning and analysis. Projected temperature requirements were compared to specifications of existing instruments to identify instrumentation development needs. Different instrument technologies, not currently employed in geotechnical instrumentation, were reviewed to identify potential improvements of existing designs for the high temperature environment. Technologies with strong potentials to improve instrument performance with relatively high reliability include graphite fiber composite materials, fiber optics, and video imagery.

As a follow-on to Sandia`s 1991 preliminary total-system performance assessment of the Yucca Mountain site, this paper presents results of some sensitivity analyses that were done using results from the 1991 study. Two conceptual models of unsaturated-zone flow and transport at Yucca Mountain were included in the study, including both aqueous and gaseous releases. The sensitivities are quite different for the two models. For the composite-porosity model, the results are most sensitive to groundwater percolation flux, gaseous transport time, container lifetime, and fuel-matrix-alteration rate. For the weeps model, the results are most sensitive to parameters used to characterize fracture flow (fracture aperture and fracture connectivity) and infiltration (percolation flux and weep-episode factor).

The event-tree method of scenario construction has been chosen for the Yucca Mountain performance assessment. Its applicability and suitability to the problem are discussed and compared with those of the Nuclear Regulatory Commission (NRC) method. The event-tree method is appropriate for an incompletely characterized site, where there must be an evolving understanding, over time, of the processes at work, for a site that may require analysis of details in specific context, and when the scenario functions to guide site characterization. Anticipating the eventual requirement for using the NRC method, we show that the event-tree method can be translated to the NRC format after final scenario screening.

Chemical vapor deposition (CVD) is a widely used method for depositing thin films of a variety of materials. Applications of CVD range from the fabrication of microelectronic devices to the deposition of protective coatings. New CVD processes are increasingly complex, with stringent requirements that make it more difficult to commercialize them in a timely fashion. However, a clear understanding of the fundamental science underlying a CVD process, as expressed through computer models, can substantially shorten the time required for reactor and process development. Research scientists at Sandia use a wide range of experimental and theoretical techniques for investigating the science of CVD. Experimental tools include optical probes for gas-phase and surface processes, a range of surface analytic techniques, molecular beam methods for gas/surface kinetics, flow visualization techniques and state-of-the-art crystal growth reactors. The theoretical strategy uses a structured approach to describe the coupled gas-phase and gas-surface chemistry, fluid dynamics, heat and mass transfer of a CVD process. The software used to describe chemical reaction mechanisms is easily adapted to codes that model a variety of reactor geometries. Carefully chosen experiments provide critical information on the chemical species, gas temperatures and flows that are necessary for model development and validation. This brochure provides basic information on Sandia`s capabilities in the physical and chemical sciences of CVD and related materials processing technologies. It contains a brief description of the major scientific and technical capabilities of the CVD staff and facilities, and a brief discussion of the approach that the staff uses to advance the scientific understanding of CVD processes.

This paper presents a brief summary of condition monitoring results from an NRC-sponsored test program on aging, condition monitoring, and accident testing of Class 1E cables. In addition, since the program has been completed and the final reports have been published, this paper presents a list of references, along with abstracts for each reference. This list can serve as a guide for finding references to any desired information on the test program in varying levels of detail. The major conclusions of the test program with regard to condition monitoring were as follows: (a) of the parameters tested, elongation at break had the best correlation with aging for the most cable types; (b) hardness and indenter modulus (performed using the cable indenter developed under Electric Power Research Institute (EPRI) sponsorship) measurements both increased with aging for some of the materials, especially the jacket materials, and the modulus measurements were significantly more sensitive to aging than the hardness measurements; (c) where indenter modulus was sensitive to aging, it was most sensitive in the later stages of aging, after the elongation had reached nearly 0%; (d) density generally increased with aging for most materials, but some changes were inconsistent; and (e) with only a few exceptions, tensile strength and a number of different electrical measurements did not correlate well with aging.

A fundamental concern in the design of the potential repository at Yucca Mountain. Nevada is the response of the host rock to the emplacement of heat-generating waste. The thermal perturbation of the rock mass has implications regarding the structural, hydrologic. and geochemical performance of the potential repository. The phenomenological coupling of many of these performance aspects makes repository thermal modeling a difficult task. For many of the more complex, coupled models, it is often necessary to reduce the geometry of the potential repository to a smeared heat-source approximation. Such simplifications have impacts on induced thermal profiles that in turn may influence other predicted responses through one- or two-way thermal couplings. The effect of waste employment layout on host-rock thermal was chosen as the primary emphasis of this study. Using a consistent set of modeling and input assumptions, far-field thermal response predictions made for discrete-source as well as plate source approximations of the repository geometry. Input values used in the simulations are consistent with a design-basis a real power density (APD) of 80 kW/acre as would be achieved assuming a 2010 emplacement start date, a levelized receipt schedule, and a limitation on available area as published in previous design studies. It was found that edge effects resulting from general repository layout have a significant influence on the shapes and extents of isothermal profiles, and should be accounted for in far-field modeling efforts.

Calculations of residual stresses in braze joints are required to validate designs for a variety of metal/ceramic joining applications. In particular, finite element analysis (FEA) codes have the capability of incorporating either elastic-plastic or minimum creep rate constitutive models for the braze material. This paper presents both elevated temperature mechanical properties correlations for the eutectic Ag-Cu alloy, along with FEA results which use this data in calculating residual stresses in a generic metal/ceramic ``shear'' type braze joint. Three constitutive relations have been developed for eutectic Ag-Cu alloy: (i) an elastic/plastic correlation incorporating temperature-dependent yield stress and work hardening data, (ii) a high temperature minimum creep rate correlation with a temperature-dependent stress exponent and (iii) a minimum creep rate correlation using the Garofalo hyperbolic sine (sinh) equation. FEA calculations are presented for a eutectic Ag-Cu braze joint between metallized alumina ceramic and either Fe-29Ni-17Co or Fe-27Ni-25Co alloys using the three different constitutive relations for the brazement. The two creep correlations, since they are time-dependent, permit a study of the effect of various cooldown cycles on the maximum residual stress in the alumina ceramic. For the cooldown profiles studied in this paper, lower residual stresses are predicted in the ceramic-relative to the elastic-plastic model - when either of the two creep models are used as the constitutive law for the eutectic Ag-Cu braze joint. A second important result is that the simulations which incorporate the Fe-29-Ni-17Co alloy show higher peak stresses than the Fe-27Ni-25Co alloy at 420°C, along with much lower peak stresses compared to Fe-27Ni-25Co alloy at room temperature. The reason for this somewhat surprising behavior can be understood in terms of the coefficient of thermal expansion for the two Fe-Ni-Co alloys.

A description of ion-irradiation-induced reduction in the photoluminescence (PL) signal from porous silicon is given and a simple model which is consistent with a nanocrystalline Si structure is presented. Ion irradiation with 250 keV Ne is used to controllably reduce the integrated PL signal by 20% after a fluence of 4*1012 Ne cm-2 and completely eliminate the PL signal after a fluence of 4*1013 Ne cm-2. The use of vacuum and air annealing to recover ion-induced damage is also described, but the high temperatures for annealing cause elimination of the PL signal.

Pyro shock loads are generated in many missile or rocket systems when stages are separated or shrouds are removed. These shocks are localized, of short rise time (10's of ns) and of very high stress level. This paper will document some anomalous behavior that occurs when pyro shock accelerometers (Endevco 7270A) are exposed to levels of high frequency shock that is higher than the manufacturer's recommended limits. Such shocks occur in many pyro shock events. Standard, accepted recording techniques can produce totally erroneous data with no obvious indicator that the data is in error. Wide band data recording, along with Fourier Analysis of the data, and dynamic analysis, made by the gage manufacturer of the transducer, allow recognition of some of the non-standard response modes excited but no method of quantifying the corrupted data has been developed. Wide band recording, which will preserve the data in the gage resonance range and above, are required to insure understanding of these pyro shock events.

We have measured the efficiency (tracks per incident neutron) of pure CR-39 for detecting DD and DT neutrons. Neutrons having average energies of 2.9 MeV (DD) and 14.8 MeV (DT) were produced by a 200-keV electrostatic accelerator and the neutron yields were measured using the associated particle counting technique. All CR-39 samples irradiated by DD or DT neutrons were etched for 2 h in a 70°, 6.25-N- NaOH bath. For bare CR-39, the efficiencies for detecting 2.9- and 14.8-MeV neutrons were found to be (1.3±0.4)×10 -4 and (5.0±1.8)×10-5, respectively. We also investigated using CR-39 and polyimide as proton radiators. For detecting 2.9-MeV neutrons, the radiators had no significant effect on efficiency; but for detecting 14.8-MeV neutrons the polyimide radiator increased the efficiency to (7.8±2.8)×10-5.

Progress in Z-pinch experiments at Sandia's Saturn facility have underscored a need for an absolute yield measurement for DD fusion neutrons. The technique chosen for making this absolute yield measurement was neutron activation of indium metal samples. To calibrate the technique, a 175-keV deuteron beam was allowed to impinge on a 3.0-μm-thick erbium deuteride target, producing neutrons through the 2H(d,n)3He fusion reaction. The neutron flux produced at 0° and incident on nominal 5-g indium samples was determined by the associated particle method. This method employed protons measured from the 2H(d,p)3H reaction to infer the neutron flux produced. After neutron irradiation, the activity of the indium samples was measured with a Ge gamma-ray detector. The total activity of the metastable state 115mIn (336.23 keV) was measured, compared with the total incident flux, and a calibration factor (indium counts/neutron/gram of indium) determined. For completeness, a calibration factor for DT neutrons from the 3H(d,n)4He fusion reaction was also obtained through the measured activity of the metastable state 114mIn(190.29 keV). The experiment and the measured calibration factors for both reactions are described in the paper.

The CR-39/range-filter technique measures ion energy by determining the maximum filter thickness which ions can penetrate. CR-39 located behind the filter records the ions. This method is used to measure peak voltage in pulsed power accelerators. We investigated range and straggling effects in this diagnostic by exposing it to 8- and 15-MeV protons for both Al and Ta filters. The range agreed with published values to better than ±6%. The range straggling decreased for higher incident ion energy and lower atomic number, as expected, although there were differences up to a factor of 1.7 between the experimental values and predictions. The dependence of the track diameter distribution on ion energy enabled us to establish a signature which is characteristic of ions which penetrate a filter, via straggling. These results can be used to evaluate the errors present when this diagnostic is used to measure accelerator voltage.

In recent proton beam experiments on PBFA-II, foam-filled gold targets and gas-filled spherical exploding pushers were shot as physics targets. Surrounding these targets were gold foils used to characterize the beam. The target fabrication and characterization are presented in this paper.

Because a direct measurement of the voltage (V) in pulsed power bremsstrahlung sources can be difficult, the energy spectrum of x rays emitted is sometimes used to infer V. Both the voltage and current in such sources vary with time. Moreover, for modern x-ray simulators with multiple cathodes, multiple voltages may exist simultaneously. We demonstrate here how such sources lead to systematic errors in several types of simple-to-field x-ray voltage measurements, especially those with broad spectral response functions, when calibrated against constant-potential bremsstrahlung spectra.

The laser-initiated LEVIS source is intended to provide an active source in the PBFA II ion diode by generation of a preformed lithium anode plasma. Visible spectra recorded with a streaked spectrograph are used to help evaluate the source performance. The absence of Stark shifting of the Li I 2p-3d and 2s-2p light confirms that a plasma is created over the anode surface. At lower LEVIS laser intensities this plasma erodes back toward the anode as the ion beam current grows. Considerable improvements in anode plasma purity have been noted when the anode is heated for several hours. The LEVIS source microdivergence estimated from the carbon ion transverse energy is less than 14 mrad, meeting the near term requirements for PBFA II target experiments. © 1992 National Technical Information Service.

Helium removal experiments were conducted in TEXTOR with a small helium self-pumping module located in a modified ALT-I limiter head. The module contained two heated nickel alloy trapping plates, a nickel deposition filament array, a Langmuir probe, flux probe, and thermocouples. The experiment examined plasma helium removal via trapping of helium ions in the deposited nickel surfaces. Such helium removal was successfully observed, with about 10% of the helium in a 10% He/D plasma being removed in a ∼1 s period. The module was found to be compatible with overall tokamak operation with essentially no sputtered nickel entering the core plasma. The temperature rise on the ion-exposed inner trapping plate, during a plasma shot, is consistent with a local sheath potential of ∼3kTe. Post-tokamak test examination of the trapping plates shows helium atom concentrations in the deposited nickel consistent with the observed helium removal, and shows very small D concentrations. © 1992 Elsevier Science Publishers B.V. All rights reserved.

We present results from initial experiments with a high impedance applied-B extraction diode on the SABRE ten stage linear induction accelerator (6 MV, 300 kA). We have demonstrated efficient coupling of power from the accelerator through an extended MTTL (Magnetically Insulated Transmission Line) into a high intensity ion beam. Both MITL electron flow in the diode region and ion diode behavior, including ion source turn-on, virtual cathode formation and evolution, enhancement delay, and ion coupling efficiency, are strongly influenced by the geometry of the diode insulating magnetic field. For our present diode electrode geometry, electrons from the cathode feed play a large role in the evolution of the virtual cathode. Both experimental data and particle-in-cell numerical simulations show that a uniform radial distribution of these feed electrons is required for uniform ion emission and efficient diode operation. © 1992 National Technical Information Service.

An active ion source, LEVIS (Laser Evaporation Ion Source), using a short pulsed Nd laser (1.06 m, 8ns) to form a thin Li vapor layer and a tuned dye laser (670.8nm, μsec) for ionization is being developed at Sandia. The LEVIS process has been developed and characterized in the laboratory. Initial experiments using these ion sources on PBFA-II produced ion beams with a low fraction of lithium ions and with low voltages. Recent experiments using improved distribution optics and anode cleaning techniques are described. © 1992 National Technical Information Service.