Finite element analysis of creep fracture initiation in a model superalloy material

Detailed finite element analyses were performed for a single edge-cracked specimen geometry under both plane stress and plane strain constraint for a superalloy material that obeys a power-law creep relationship. The objectives of these analyses were to elucidate the stationary creep crack-tip fields and to provide guidance for the experimental measurement of crack-tip deformations. New results demonstrate that, for both plane stress and plane strain, the angular variations in the creep strain fields do not agree with HRR-type predictions, although the radial variations are in agreement with HRR-type creep strain field predictions in a zone very near the crack tip. Thus, the use of experimental measurement of surface displacement and/or strain data for the location of HRR-type fields may not be possible, unless modifications to the existing HRR-type theory are made. It is also noted that the size of the stress-based HRR-dominance zone is only a fraction of the creep zone size in plane stress, and is very small (especially along =0°) compared to the creep zone size in plane strain. Furthermore, the dominance of the singular strain fields are at least two orders of magnitude smaller than the corresponding stress dominance zones. As such, unless the microstructural features of the material are smaller than the dimensions of the dominance zones, the basis for using stress or strain-based fracture parameters derived from the HRR-type fields for prediction of creep fracture initiation is unclear.     

The lactulose breath hydrogen test and small intestinal bacterial overgrowth

OBJECTIVES: To i) document the sensitivity and specificity of a combined scintigraphic/lactulose breath hydrogen test for small intestinal bacterial overgrowth and ii) investigate the validity of currently accepted definitions of an abnormal lactulose breath hydrogen test based on "double peaks" in breath hydrogen concentrations. METHODS: Twenty-eight subjects were investigated with culture of proximal small intestinal aspirate and a 10-g lactulose breath hydrogen test combined with scintigraphy. Gastroduodenal pH, the presence or absence of gastric bacterial overgrowth, and the in vitro capability of overgrowth flora to ferment lactulose were determined. RESULTS: Sensitivity (16.7%) and specificity (70.0%) of the lactulose breath hydrogen test alone for small intestinal bacterial overgrowth were poor. Combination with scintigraphy resulted in 100% specificity, because double peaks in serial breath hydrogen concentrations may occur as a result of lactulose fermentation by cecal bacteria. Sensitivity increased to 38.9% with scintigraphy, because a single rise in breath hydrogen concentrations, commencing before the test meal reaches the cecum, may occur in this disorder. Sensitivity remained suboptimal irrespective of the definition of small intestinal bacterial overgrowth used, the nature of the overgrowth flora, favorable luminal pH, the presence of concurrent gastric bacterial overgrowth, or the in vitro ability of the overgrowth flora to ferment lactulose. CONCLUSIONS: Definitions of an abnormal lactulose breath hydrogen test based on the occurrence of double peaks in breath hydrogen concentrations are inappropriate. Not even the addition of scintigraphy renders this test a clinically useful alternative to culture of aspirate for diagnosing small intestinal bacterial overgrowth.     

Legal issues in sharing the benefits of biodiversity prospecting

The National Cancer Institute (NCI) is the US Government's principal agency for research on the prevention, diagnosis and treatment of cancer. A critical component of the Institute's mission is the identification and development of new and promising treatments for cancer and AIDS. For many years these efforts have included a program to investigate natural products for potential new therapeutic agents. In 1986, with the advent of new screening techniques, the National Cancer Institute stepped up its exploration of natural products and began world-wide collections of plants in tropical and subtropical regions. In recognition of the principles of the Biodiversity Treaty, NCI appreciates that continued access to the natural products of these countries depends on the Institute's ability to recognize the contributions of these source countries and their indigenous peoples, and to provide them adequate incentives to conserve their natural resources for the purposes of drug discovery. Accomplishing this goal presented several legal issues for the National Cancer Institute. As an agency of the US government, the NCI has an adjunct statutory mission to facilitate the transfer of technology developed through the Institute's programs into the private sector for further development and commercialization, and NCI operates under a national policy to use the patent system to transfer Federally supported research to the private domestic sector. Reliance on patent law may limit the Institute's ability to recognize the rights of source countries and their indigenous people and provide compensation for their contributions. However, other legal instruments, such as contracts, can serve as interim measures to provide compensation to source countries and indigenous populations. The National Cancer Institute's Letter of Collection agreement (LOC, formerly the "Letter of Intent'), is an example of an alternative means that "fills-in the gaps' created by patent law and through which source countries may share in the benefits of natural product development.     

Piecewise training for structured prediction

A drawback of structured prediction methods is that parameter estimation requires repeated inference, which is intractable for general structures. In this paper, we present an approximate training algorithm called piecewise training (PW) that divides the factors into tractable subgraphs, which we call pieces, that are trained independently. Piecewise training can be interpreted as approximating the exact likelihood using belief propagation, and different ways of making this interpretation yield different insights into the method. We also present an extension to piecewise training, called piecewise pseudolikelihood (PWPL), designed for when variables have large cardinality. On several real-world natural language processing tasks, piecewise training performs superior to Besag’s pseudolikelihood and sometimes comparably to exact maximum likelihood. In addition, PWPL performs similarly to PW and superior to standard pseudolikelihood, but is five to ten times more computationally efficient than batch maximum likelihood training.     

Linguistic Problems with Requirements and Knowledge Elicitation

Human and conversational aspects of requirements and knowledge identification are employed to show that requirements ‘engineering’ is not the same as civil engineering or scientific problem solving. Not only can requirements not be made fully explicit at the start of a project, they cannot be made fully explicit at all. A need is identified to enhance computer-based information systems (CBIS) development methods to accommodate: plurality of incommensurable perspectives, languages and agendas; dynamic representations of system features that can be experienced rather than abstracted and forced into an abstract paper-based representation; recognition that CBIS development is in general a continuous process where users changing their minds is a natural and necessary indication or organisational vitality.  It is suggested that prototyping and rapid application development go some way to addressing these requirements but that they require further development in the light of the theoretical light thrown on the nature of the problem.     

Failure of triiodothyronine to inhibit TSH-mediated thyroid hormone release in man

We studied the effect of short-term triiodothyronine administration on thyroid gland responsivity to exogenous thyrotropin in four euthyroid human subjects. Thyroidal iodine release and serum thyroxine during daily im injections of bovine TSH were not significantly inhibited, despite a four-fold elevation in serum T3 concentrations. This negative finding contrasts with earlier positive reports of a regulatory "short-loop" effect of elevated circulating T3 on the thyroid gland. This difference may be due either to the use in previous murine or in vitro studies of non-physiologic, high doses of exogenous T3, or failure to control the withdrawal of the trophic effect of endogenous TSH in man on the subsequent glandular response.     

A Raman microprobe study of electrical treeing in polyethylene

The chemical nature of electrical tree growth in a blend of high and low density polyethylene has been studied by confocal Raman microprobe spectroscopy. The observed spectra, which are easily perturbed by the Raman probe beam, can be described in terms of three components, the relative intensities of which vary from place-to-place on the sample. Throughout the body of the tree, the usual Raman bands of polyethylene are seen, but superimposed upon a pronounced fluorescent background. This suggests a degree of material degradation throughout the structure, even where there is no visible evidence of electrical discharge damage or ageing. The individual channels that make up the fractal structure of the tree can then be divided into two distinct categories. Within the core of the tree, their Raman spectra are made up of two elements; fluorescence, plus the G and D bands of sp² hybridized carbon. Here, the tree channels are best thought of as hollow tubules surrounded by a carbonaceous shell. At the tips, the channels are characterized simply by the Raman bands of polyethylene superimposed upon a reduced fluorescent background, irrespective of their size. That is, in this region, the tree channels are simply hollow tubules within the dielectric. The transition between these two structures occurs relatively abruptly. These observations are related to the treeing process.     

Grain Boundary Responses to Heterogeneous Deformation in Tantalum Polycrystals

The evolution of heterogeneous deformation in a tantalum polycrystal was examined during a three-point bending experiment using electron backscatter pattern mapping. Slip bands formed at strains as low as 1%, and they became more intense with strain. Heterogeneous deformation was evident as intragranular orientation gradients as large as 30° were observed after a strain of about 8%. Nonmonotonic changes in the local average misorientation distribution were observed, implying that dislocation substructure developed in a complex manner. Slip bands were analyzed using plane traces computed from local orientation information. With the assumption of uniaxial stress, Schmid factors for favorable slip systems were identified for each grain and compared with observations, showing evidence for macroscopic activity on both {110} and {112} slip systems. Reconstructed boundary data were used to estimate the geometric potential for slip transfer at grain boundaries. The correlations indicated that when active slip systems were favorably oriented for slip transfer across the boundary, it was often observed in the form of continuous slip bands aligned across the boundary. In boundaries where geometrical alignment and Schmid factors were not favorable for slip transfer, there was a higher likelihood to form ledges (topographic discontinuities) along the grain boundaries. Dislocation pileups at grain boundaries were also correlated with a low potential for slip transfer.     

Investigation of NCN and prompt-NO formation in low-pressure C1–C4 alkane flames

Temperature, CH, NCN, and NO profiles were measured for eight low-pressure hydrocarbon flames fueled by methane, ethane, propane, and butane using laser-induced fluorescence (LIF) diagnostics. These measurements were used (1) to assess NCN and prompt-NO formation chemistry across a series of fuels of increasing number of carbons at different equivalence ratios (? = 1.07 and 1.28); (2) to examine the predictive capabilities of current C1–C4 hydrocarbon and NCN formation/consumption combustion mechanisms on properly capturing prompt-NO formation and (3) to examine the postulation that additional prompt-NO precursors (other than CH) exist for fuels larger than methane. For a given equivalence ratio, the measured peak CH concentration is fairly constant across all four fuels, while both the peak NCN and post-flame NO concentrations steadily increase. Furthermore, it is found that as the fuels increase in number of carbons, i.e., methane to butane, the correlation between the peak NCN and post-flame NO remains high, while the correlation between peak CH and peak NCN and peak CH and post-flame NO becomes increasingly lower. This is especially evident for rich flame cases. The experimental profiles are compared to numerical calculations using two comprehensive kinetic mechanisms suitable for C4 chemistry, where the CH + N₂ → NCN + H reaction is assumed as the only prompt-NO initiation reaction. For the ? = 1.28 flame cases, CH is over-predicted using both mechanisms for all four fuels and by as much as 60%, while for the ? = 1.07 cases, CH is predicted to within 15% of the experimentally-derived results, although there is some discrepancy concerning the spatial locations of the CH profiles. For both NCN and NO, there is an increasing under-prediction for the ? = 1.28 cases as the fuel increases in number of carbons, while for the ? = 1.07 cases there is a systematic under-prediction of NCN and NO with a weaker (although evident) fuel dependence. From the experimental results and the comparison to modeling predictions, it is apparent that additional work concerning CH formation and consumption kinetics is necessary to accurately capture the CH concentration profiles across a broad range of conditions. Furthermore the comparisons to the modeling predictions using only a single prompt-NO precursor, CH, indicate a reasonable plausibility that (an) additional prompt-NO precursor (s) exist and become important when considering fuels larger than methane, especially under rich flame conditions. Possible precursors in addition to CH are discussed.