This paper extends previous results on nonlinear Schwarz preconditioning ([4]) to unstructured finite element elliptic problems exploiting now nonlocal (but small) subspaces. The non-local finite element subspaces are associated with subdomains obtained from a non-overlapping element partitioning of the original set of elements and are coarse outside the prescribed element subdomain. The coarsening is based on a modification of the agglomeration based AMGe method proposed in [8]. Then, the algebraic construction from [9] of the corresponding non-linear finite element subproblems is applied to generate the subspace based nonlinear preconditioner. The overall nonlinearly preconditioned problem is solved by an inexact Newton method. Numerical illustration is also provided.

The activity concentration of Cesium-137 ({sup 137}Cs) and naturally-occurring Polonium-210 ({sup 210}Po) were measured in the muscle tissue, kidney and liver of Pacific walrus (Odobenus rosmarus divergens) and bearded seal (Erignathus barbatus) collected by native hunters from the Bering Sea. The mean {sup 137}Cs concentrations in muscle, liver and kidney of Pacific walrus were 0.07, 0.09 and 0.07 Bq kg{sup -1} (N= 5, wet weight), respectively, and 0.17, 0.10, and 0.17 Bq kg{sup -1} (N=2, wet weight), respectively, in bearded seal. In general, {sup 137}Cs tissue concentrations are significantly lower than those previously reported for mammals from other regions. By comparison, {sup 210}Po activity concentrations appear to be higher than those reported elsewhere but a larger variation. The mean {sup 210}Po concentration in the muscle tissue, liver and kidney of Pacific walrus (N=5, wet weight) were 28.7, 189, and 174 Bq kg{sup -1}, respectively. This compares with {sup 210}Po concentration values (N=2, wet weight) of 27, 207, and 68 Bq kg{sup -1} measured in the muscle tissue, liver and kidney, of bearded seal, respectively. Estimated bioaccumulation factors--as defined by the radionuclide concentration ratio between the target tissue to that in sea water--were two to three orders of magnitude higher for …

Worldwide data on terrorist incidents between 1968 and 2004 gathered by the RAND corporation and the Oklahoma City National Memorial Institute for the Prevention of Terrorism (MIPT) were assessed for patterns and trends in morbidity/mortality. Adjusted data analyzed involve a total of 19,828 events, 7,401 ''adverse'' events (each causing {ge}1 victim), and 86,568 ''casualties'' (injuries) of which 25,408 were fatal. Most terror-related adverse events, casualties and deaths involved bombs and guns. Weapon-specific patterns and terror-related risk levels in Israel (IS) have differed markedly from those of all other regions combined (OR). IS had a fatal fraction of casualties about half that of OR, but has experienced relatively constant lifetime terror-related casualty risks on the order of 0.5%--a level 2 to 3 orders of magnitude more than those experienced in OR that increased {approx}100-fold over the same period. Individual event fatality has increased steadily, the median increasing from 14 to 50%. Lorenz curves obtained indicate substantial dispersion among victim/event rates: about half of all victims were caused by the top 2.5% (or 10%) of harm-ranked events in OR (or IS). Extreme values of victim/event rates were approximated fairly well by generalized Pareto models (typically used to fit to data on forest …

Factors affecting sensing of small quantities of fissionable material in large sea-going cargo containers by neutron interrogation and detection of {beta}-delayed photons are explored. The propagation of variable-energy neutrons in cargos, subsequent fission of hidden nuclear material and production of the {beta}-delayed photons, and the propagation of these photons to an external detector are considered explicitly. Detailed results of Monte Carlo simulations of these stages in representative cargos are presented. Analytical models are developed both as a basis for a quantitative understanding of the interrogation process and as a tool to allow ready extrapolation of the results to cases not specifically considered here.

We have developed processes for producing ultra-smooth nanoscale programmed substrate defects that have applications in areas such as thin film growth, EUV lithography, and defect inspection. Particle, line, pit, and scratch defects on the substrates between 40 and 140 nm wide 50 to 90 nm high have been successfully produced using e-beam lithograpy and plasma etching in both Silicon and Hydrosilsequioxane films. These programmed defect substrates have several advantages over those produced previously using gold nanoparticles or polystyrene latex spheres--most notably, the ability to precisely locate features and produce recessed as well as bump type features in ultra-smooth films. These programmed defects were used to develop techniques for film defect mitigation and results are discussed.

We present first-principles total-energy density-functional theory electronic structure calculations for the neutral and charge states of an oxygen vacancy in KH{sub 2}PO{sub 4} (KDP). Even though the overall DOS profiles for the defective KDP are quite similar to those of the perfect KDP, the oxygen vacancy in the neutral and +1 charge states induces defect states in the band gap. For the neutral oxygen vacancy, the gap states are occupied by two electrons. The difference between the integral of the total density of states (DOS) and the sum of the DOS projected on the atoms of 0.98 |e|, indicates that one of the two electrons resulting from the removal of the oxygen atom is trapped in the vacancy, while the other tends to delocalize in the neighboring atoms. For the +1 charge oxygen vacancy, the addition of the hole reduces the occupation of the filled gap-states in the neutral case from two to one electron and produces new empty states in the gap. The new empty gap states are very close to the highest occupied states, leading to a dramatic decrease of the band gap. The difference between the integral of the total DOS and the sum of the DOS …

Nanoscale materials provide unique properties that will enable new technologies and enhance older ones. One area of intense activity in which nanoscale materials are being used is in the development of new functional materials for battery applications. This effort promises superior materials with properties that circumvent many of the problems associated with traditional battery materials. Previously we have worked on several approaches for using nanoscale materials for application as cathode materials in rechargeable Li batteries. Our recent work has focused on synthesizing MnO2 nanoparticles and using these in layer-by-layer (LbL) structures to probe the redox properties of the nanoparticles. We show that the aqueous colloidal nanoparticles produced by butanol reduction of tetramethylammonium permanganate can be trapped in thin films using a layer-by-layer deposition approach, and that these films are both redox active and exhibit kinetically facile electrochemical responses. We show cyclic voltammetry of MnO2 colloidal nanoparticles entrapped in a LbL thin film at an ITO electrode surface using poly(diallyldimethylammonium chloride) (PDDA). CV experiments demonstrate that Li+ insertion accompanies Mn(IV) reduction in LiClO4 supporting electrolytes, and that reduction is hindered in supporting electrolytes containing only tetrabutylammonium cations. We also show that electron propagation through multilayer films is facile, suggesting that electrons …

The molecular structures of drug target proteins and receptors form the basis for 'rational' or structure guided drug design. The majority of target structures are experimentally determined by protein X-ray crystallography, which as evolved into a highly automated, high throughput drug discovery and screening tool. Process automation has accelerated tasks from parallel protein expression, fully automated crystallization, and rapid data collection to highly efficient structure determination methods. A thoroughly designed automation technology platform supported by a powerful informatics infrastructure forms the basis for optimal workflow implementation and the data mining and analysis tools to generate new leads from experimental protein drug target structures.

We present experimental results in three areas. Solar cells with an efficiency of 19% have been fabricated with an absorber bandgap in the range of 1.1-1.2 eV. Properties of solar cells fabricated with and without an undoped ZnO layer were compared. The data show that high efficiency cells can be fabricated without using the high-resistivity or undoped ZnO layer. Properties of CIGS solar cells were fabricated from thin absorbers (1 {micro}m) deposited by the three-stage process and simultaneous co-deposition of all the elements. In both cases, solar cells with efficiencies of 16%-17% are obtained.

We report the deposition of improved hydrogenated amorphous silicon germanium (a-SiGe:H) films by the hot wire CVD (HWCVD) technique using a tantalum filament operating at a low temperature. We gauge the material quality of the a-SiGe:H films by comparing infrared, small angle X-ray scattering (SAXS), photocapacitance, and conductivity measurements to earlier results, and fabricate single-junction n-i-p solar cell devices using these i-layers.

By analogy to the different accretion states observed in black-hole X-ray binaries (BHXBs), it appears plausible that accretion disks in active galactic nuclei (AGN) undergo a state transition between a radiatively efficient and inefficient accretion flow. If the radiative efficiency changes at some critical accretion rate, there will be a change in the distribution of black hole masses and bolometric luminosities at the corresponding transition luminosity. To test this prediction, the author considers the joint distribution of AGN black hole masses and bolometric luminosities for a sample taken from the literature. The small number of objects with low Eddington-scaled accretion rates m < 0.01 and black hole masses M{sub BH} < 10{sup 9} M{sub {circle_dot}} constitutes tentative evidence for the existence of such a transition in AGN. Selection effects, in particular those associated with flux-limited samples, systematically exclude objects in particular regions of the (M{sub BH}, L{sub bol}) plane. Therefore, they require particular attention in the analysis of distributions of black hole mass, bolometric luminosity, and derived quantities like the accretion rate. The author suggests further observational tests of the BHXB-AGN unification scheme which are based on the jet domination of the energy output of BHXBs in the hard state, …

Better to understand the status of fusion research in the year 2003 we will first put the research in its historical context. Fusion power research, now beginning its sixth decade of continuous effort, is unique in the field of scientific research. Unique in its mixture of pure and applied research, unique in its long-term goal and its promise for the future, and unique in the degree that it has been guided and constrained by national and international governmental policy. Though fusion research's goal has from the start been precisely defined, namely, to obtain a net release of energy from controlled nuclear fusion reactions between light isotopes (in particular those of hydrogen and helium) the difficulty of the problem has spawned in the past a very wide variety of approaches to the problem. Some of these approaches have had massive international support for decades, some have been pursued only at a ''shoestring'' level by dedicated groups in small research laboratories or universities. In discussing the historical and present status of fusion research the implications of there being two distinctly different approaches to achieving net fusion power should be pointed out. The first, and oldest, approach is the use of strong magnetic …

The Red-Sequence Cluster Survey (RCS) provides a large and deep photometric catalog of galaxies in the z' and R{sub c} bands for 90 square degrees of sky, and supplemental V and B data have been obtained for 33.6 deg{sup 2}. They compile a photometric redshift catalog from these 4-band data by utilizing the empirical quadratic polynomial photometric redshift fitting technique in combination with CNOC2 and GOODS/HDF-N redshift data. The training set includes 4924 spectral redshifts. The resulting catalog contains more than one million galaxies with photometric redshifts < 1.5 and R{sub c} < 24, giving an rms scatter {delta}({Delta}z) < 0.06 within the redshift range 0.2 < z < 0.5 and {sigma}({Delta}z) < 0.11 for galaxies at 0.0 < z < 1.5. They describe the empirical quadratic polynomial photometric redshift fitting technique which they use to determine the relation between red-shift and photometry. A kd-tree algorithm is used to divide up the sample to improve the accuracy of the catalog. They also present a method for estimating the photometric redshift error for individual galaxies. They show that the redshift distribution of the sample is in excellent agreement with smaller and much deeper photometric and spectroscopic redshift surveys.

None

This paper presents a personal account of the scientific and professional adventures of Bikash Sinha on the occasion of the celebration of his 60th birthday held in Calcutta on Feb 7, 2005.

Photoelectron spectroscopy was used to determine the compositional and electronic changes occurring in Cu(In,Ga)Se2 thin films as a result of immersion in aqueous ammonia solution. We find that NH4OH-treated CIGS surfaces are preferentially etched of indium and gallium, resulting in the formation of a thin layer of a degenerate Cu-Se compound that we tentatively identify as Cu2Se. The work function of ammonia-treated samples is found to increase by 0.6 eV relative to as-grown CIGS thin films. The uniformity of chemical bath effects (etching & deposition) was found to be improved by the addition to the bath of a non-ionic surfactant. Initial device results show that the new surfactant-based chemical bath deposition (CBD) method may lead to better and thinner CdS buffer layers.

GaInNAs, potentially useful in a 4-junction GaInP2/GaAs/GaInNAs/Ge solar cell, suffers from very low minority-carrier collection lengths. To date, the currents available from GaInNAs solar cells are not high enough to increase the efficiency of a 3-junction device by adding this fourth junction. Here, we grow p-i-n GaInNAs solar cells by molecular-beam epitaxy with wide, intrinsic base layers and internal quantum efficiencies near 1.0. If similar 1.0-eV GaInNAs junctions can be successfully integrated into the 3-junction structure, the resulting 4-junction cell would have a higher efficiency.

Measurement of recombination and minority-carrier lifetimes has become a central activity in photovoltaic technology. The primary measurement techniques for silicon technologies are based on photoconductive decay (PCD) and microwave reflectance (mPCD). The measurement of the correct recombination lifetime depends on the carriers being confined to a given spatial region of a diagnostic structure. The internal electric fields separate the charges, and the measured value does not represent the real minority-carrier lifetime. In these cases, the measured quantity is a function of the true lifetime and the sample structure. Here, we examine these effects, both experimentally and theoretically, for the n+-p device structure common to terrestrial photovoltaics.

A Service Class Description (SCD) is an effective meta-data based approach for discovering Deep Web sources whose data exhibit some regular patterns. However, it is tedious and error prone to create an SCD description manually. Moreover, a manually created SCD is not adaptive to the frequent changes of Web sources. It requires its creator to identify all the possible input and output types of a service a priori. In many domains, it is impossible to exhaustively list all the possible input and output data types of a source in advance. In this paper, we describe machine learning approaches for automatic generation of the data types of an SCD. We propose two different approaches for learning data types of a class of Web sources. The Brute-Force Learner is able to generate data types that can achieve high recall, but with low precision. The Clustering-based Learner generates data types that have a high precision rate, but with a lower recall rate. We demonstrate the feasibility of these two learning-based solutions for automatic generation of data types for citation Web sources and presented a quantitative evaluation of these two solutions.

The production of supersonic jets of material via the interaction of a strong shock wave with a spatially localized density perturbation is a common feature of inertial confinement fusion and astrophysics. The behavior of two-dimensional (2D) supersonic jets has previously been investigated in detail [J. M. Foster et. al, Phys. Plasmas 9, 2251 (2002)]. In three-dimensions (3D), however, there are new aspects to the behavior of supersonic jets in compressible media. In this paper, the commissioning activities on the National Ignition Facility (NIF) [J. A. Paisner et al., Laser Focus World 30, 75 (1994)] to enable hydrodynamic experiments will be presented as well as the results from the first series of hydrodynamic experiments. In these experiments, two of the first four beams of NIF are used to drive a 40 Mbar shock wave into millimeter scale aluminum targets backed by 100 mg/cc carbon aerogel foam. The remaining beams are delayed in time and are used to provide a point-projection x-ray backlighter source for diagnosing the three-dimensional structure of the jet evolution resulting from a variety of 2D and 3D features. Comparisons between data and simulations using several codes will be presented.

None

None

We report on a direct measurement of two-dimensional potential distribution on the surface of Cu(In,Ga)Se2 thin films using a nanoscale electrical characterization of scanning Kelvin probe microscopy both in air and in ultra-high vacuum. The potential measurement reveals a higher surface potential or a smaller work function on grain boundaries (GBs) of the film than on the grain surfaces. This demonstrates the existence of a local built-in potential on GBs, and the GB is positively charged. The role of the built-in potential in device performance was further examined and found to be positive, by tuning Ga content or bandgap of the film. With increasing Ga content, the potential drops sharply in a Ga range of 28%-38%. Comparing the change in the built-in potential to the theoretical and experimental photoconversion efficiencies, we conclude that the potential plays a significant role in the device conversion efficiency of NREL's three-stage Cu(In,Ga)Se2 device.

Magnetohydrodynamics (MHD) turbulence theory, often employed satisfactorily in astrophysical applications, has often focused on parameter ranges that imply nearly equal values of kinetic and magnetic energies and length scales. However, MHD flow may have disparity magnetic Prandtl number, dissimilar kinetic and magnetic Reynolds number, different kinetic and magnetic outer length scales, and strong anisotropy. Here a phenomenology for such ''non-equipartitioned'' MHD flow is discussed. Two conditions are proposed for a MHD flow to transition to strong turbulent flow, extensions of (1) Taylor's constant flux in an inertial range, and (2) Kolmogorov's scale separation between the large and small scale boundaries of an inertial range. For this analysis, the detailed information on turbulence structure is not needed. These two conditions for MHD transition are expected to provide consistent predictions and should be applicable to anisotropic MHD flows, after the length scales are replaced by their corresponding perpendicular components. Second, it is stressed that the dynamics and anisotropy of MHD fluctuations is controlled by the relative strength between the straining effects between eddies of similar size and the sweeping action by the large-eddies, or propagation effect of the large-scale magnetic fields, on the small scales, and analysis of this balance in principle …