79717 research outputs found
Sort by
Parallelizing WHILE loops
Get PDFTwo methods for parallelizing WHILE loops are presented. The first method converts a WHILE loop into a FORALL construct, and the second method pipelines a WHILE loop. Each of the methods is based on a transformation that makes explicit the loop counting. Also, we propose two parallel WHILE constructs
High-voltage characteristics of twin bundle drake, cardinal, and curlew with bundle spaces and armor grip suspension units
No full textCharacterization and Investigations of Thin-Film Materials with X-ray Photoelectron Spectroscopy
Get PDFSurface characterization of materials is widely utilized over a range of disciplines and essential to the development of new materials, technology, or processes. Metal oxide nanoclusters have shown promise as potential new generation photoresist materials for extreme ultraviolet (EUV) nanolithography. Organotin clusters have been proposed as potential candidates for resist materials due to a high photoabsorption cross section in the EUV energy range. Before industrial implementation, these new materials must undergo rigorous analysis through the use of numerous characterization techniques to verify effectiveness and satisfactory performance. In addition to characterization, not much information is known about the radiation induced mechanism that causes a solubility transition; a key component for acting as a photoresist. The use of near ambient pressure X-ray photoelectron spectroscopy (NAPXPS) provides the ability to study chemical changes in organotin nanocluster thin films during radiation exposure in a range of ambient environments. NAPXPS using synchrotron X-rays determined that impinging photon energy can play a role in the solubility transition mechanism since the total electron yield is dependent on the photon energy. The presence of ambient oxygen was also shown to enhance resist sensitivity. NAPXPS with a monochromated Al Kα X-rays were used to measure a contrast curve, further highlighting oxygen’s ability to enhance resist sensitivity, while also showing a decrease in sensitivity for ambients of nitrogen, water, and methanol. Thermal NAPXPS studies following the solubility transition in the resist determined a significant amount of carbon remains in the film even though carbon removal was hypothesized to be a primary step during solubility transition. This led to the conclusion that a metal oxide polymer is formed following sufficient X-ray exposure and annealing. Metal oxide materials have also shown promise as oxidation catalysts. The conversion of volatile organic compounds to non-toxic molecules like CO2 is important for pollution control. Interactions of near ambient pressures of 2-propanol (IPA) with a well ordered SnO2 surface has yet to be studied from a mechanistic standpoint. For these studies, a SnO2 single crystal was prepared with a stoichiometric oxidized surface and characterized with low energy electron diffraction (LEED) and valence band spectra. NAPXPS was used to track chemical changes in the Sn surface oxidation state and adsorbate reactions during exposure to up to 3 mbar of IPA and mixtures of IPA and oxygen at 400, 500, and 600 K. The reaction products were measured using mass spectrometry to compliment the NAPXPS results. Oxygen was found to be required for the complete conversion of IPA to CO2 to prevent the surface reduction of the Sn, which otherwise would yield the intermediate product acetone. Ultimately, surface characterization is imperative towards forming foundational chemical knowledge of materials, which can lead to new and improved technology
Creative Placemaking as a Composition Pedagogy
Get PDFThis thesis explores the ways in which creative placemaking, a neighborhood-based practice for building community, can offer community-building and civic action wisdom as a model for composition. This model brings attention to spatial metaphors for rhetoric and teaching that have persisted for millennia; it re-focuses us on community; it encourages a reconsideration of what, exactly, citizenship could mean by connecting it to dwelling; and it unites that dwelling and citizenship with creative, collaborative invention. The purpose of this thesis is to offer a preliminary model for these practices. It does this first by examining how citizenship has been understood as an aim for rhetorical training. Then, this thesis suggests ways in which dwelling can be encouraged in the classroom. This dwelling is linked to invention both in composition and in invention of creative placemaking. Invention is considered for its social realm, relating to the collaboration of creative placemaking, and how the inventing process can be creative. Finally, this thesis leads to a model grounded in topoi by using an adapted version of Aristotle's Common Topics as a way to guide students' thinking about issues in their communities. This is written by someone who teaches First Year Writing to others who teach First Year Writing to offer more tools for helping students think and write critically, inclusively, and hopefully about changing their communities
The Role of Reactivity: Experiences in Clinical Decision-Making and Countertransference of Expert Trauma Counselors
Get PDFThis dissertation presents two qualitative studies investigating expert trauma counselors’ experiences of clinical decision-making and countertransference. Chapter 2, Acutely Alert: A Grounded Theory of Clinical Decision-Making Processes of Expert Trauma Counselors addresses the research question, what are the clinical decision-making experiences and processes of expert trauma counselors? Chapter 3, Exploring Self: A Grounded Theory of Expert Trauma Counselors Encountering Countertransference examines, how do expert trauma counselors perceive, experience, and use countertransference? A purposeful sample of ten expert counselors participated in three rounds of semi-structured interviews. Data collected during interviews were transcribed, analyzed and used to develop subsequent interview questions, continuing until conceptual saturation occurred. Data analysis procedures included open, axial, and selective coding (Corbin & Strauss, 2014). The study utilized measures to promote trustworthiness delineated by Lincoln & Guba (1985) and Morrow (2005). Chapter 2 findings included three categories with a central category, evaluating reactivity. They comprise a theory representing how participants’ trauma-specific heightened awareness and accumulated knowledge inform clinical decision-making. Chapter 3 findings revealed four categories with a central category, experiential analysis, which details participants’ analysis of their reactivity. The resulting theory illustrates a process of self-inquiry as participants encounter client dysregulation. Implications for practitioners, supervisors, and counselor educators include examination of one’s own reactivity and heightened awareness and the impact on treatment and the working alliance, establishing one’s own definition and understanding of countertransference and clinical intuition, and developing a clinical routine of reflective practice and countertransference management strategies
Application and Validation of Geant4 Modeling for Optimization of Complex Structures for Fast Neutron Detection
No full textFast neutron detection is important to the detection of illicit nuclear materials and can help prevent those materials from being used or transported without discovery. Previous research has shown that Heavy Oxide Inorganic Scintillator crystals can be used to detect fast neutrons by capitalizing on their high inelastic scattering cross section. Composite scintillator designs combining crushed Gadolinium Ortho-silicate and plastic light guides have been proposed for enhanced fast neutron detection based on similar heavy inorganic crystals. In this thesis we have performed laboratory measurements and devised a Geant4 simulation of the ZEBRA design and several referenced detectors in order to characterize the neutron and photon detection abilities of ZEBRA detectors. We have validated the simulation and demonstrated its use to further optimize parameters for this promising detector technology
An Investigation into the Production of Longer Hydrocarbons, Synthesis Gas, and Ammonia via a Nonthermal Plasma Microreactor
No full textNonthermal electrical plasmas can allow energy intensive reactions to occur without the use of high operating temperatures or pressures. This is because the collisions of electrons accelerated through an electric field with molecules is the cause of reactions as opposed to thermal collisions between molecules. In this study, the production of larger hydrocarbons and synthesis gas from methane and the production of ammonia from nitrogen and hydrogen were investigated experimentally and then empirically modeled. It was found that selectivity and energy efficiency of methane dry reforming was greatly impacted by increasing the pressure of the reactor.
Carbon dioxide, air, or a combination of the two had to be added to the methane mixture in order to mitigate carbon deposition within the reactor. Methods to mitigate or eliminate deposition were examined with longevity studies to understand how long-term operation of the reactor would affect product distribution and discharge stability.
Ammonia production was possible in the electrical plasma despite the high bond energy of nitrogen. Product yields were much lower when compared to methane reforming. It was found that electron current was much more important for this process compared to composition, confirming that the electrons are the main driver of this reactor.
A microreactor was used because shorter discharges require less voltage, allowing reactions to occur with a relatively small voltage (~500 V) compared to reactors with larger distances between electrodes. This microreactor had the same geometry for all experiments. This allowed direct comparisons between the methane reaction system and the production of ammonia, which gave insights into how these reactions proceed in the plasma.
From these studies, it is possible there may be future opportunities to tune the product distribution in methane dry reforming. With a higher energy efficiency at higher pressures, plasma processes may be made more economically viable. Conversion and selectivity in the plasma can be held constant over longer time periods if efforts to mitigate deposition in the reactor and on the electrodes are taken. While it was shown that ammonia could be produced in the plasma microreactor at ambient pressure, improvements will be need to be made to improve the efficiency of the process to make it more economically viable
Thermal and Aerodynamic Effects of Surface Roughness Patterns on Additive Manufactured Heat Sinks
Get PDFBreakthroughs in heat sink design are now limited by conventional manufacturing techniques which can only produce basic fin shapes with homogenous surface structuring. Additive Manufacturing (AM) is a potential disruptive technology for heat sink design and other convective heat transfer solutions, as surface properties can be directionally varied to maximize performance for a prescribed environment and flow structure. This study examines the thermal and aerodynamic performance effects of surface roughness variation along the height and length of plane fin heat sinks. Pyramid structures have been selectively applied to the surface of the sink to produce uniform surface roughnesses of 5, 15, 30, 43 and 71 microns on heat sinks that have been created to have a roughness gradient in directions parallel and perpendicular to ambient flow. In anticipation of applicability to high air speed aerospace environments, the experimental facility is placed in the center of a wind tunnel in order to expose the sink to the maximum possible velocity, which causes the flow structure to be considered an open flow. A plane flat surface and a heat sink with a uniform surface roughness of 5 microns, validated by a white light interferometer, are compared to theoretical expectations found using Nusselt number correlations for turbulent flow over a flat plate for the flat surface, as well as the Gnielinski correlation and the Colburn equation for the heat sinks. A thermal conductivity of 11.9 W/m-K is found using a Thermal Interface Material (TIM) tester, showing a significant reduction in this property relative to the expected property of bulk aluminum which has been conventionally machined. This reduction is due primarily to the porosity of the internal structuring which is an artifact of the Selective Laser Melting (SLM) process used to fabricate the heat sinks. Using a 7-plane fin heat sink where each 2.5 mm (0.1 in.) thick fin is spaced 3.3 mm (0.13 in.) apart, a minimum thermal resistance of 0.23 K/W is found. Graded roughness in the direction of the flow is found to have no effect on thermal resistance, however, gradients in the direction normal to the bulk flow are found to decrease the thermal resistance by 3.5% for a Reynolds number range of 8,300 to 28,700 where the Reynolds number is based on the characteristic length of double the spacing between the fins. Improvements in thermal resistance reach as high as 15.2% when comparing heat sink with full maximum surface roughness patterns distributed evenly across the fins, coming at a weight increase of 4-12%. Intelligent heat sink design which exploits the surrounding flow conditions can improve performance drastically with minimal increase to heat sink weight and volume. However, this functionally graded surface roughness parameter is only possible through AM techniques. Positive impacts on performance from surface roughness gradients introduces the possibility of finding AM surface roughness patterns which are tuned using numerical modelling of the surrounding flow structure and hydrodynamic boundary layer development
Power Network Parameter Correction via Sparse Unsupervised Regression
Get PDFAccurate information of power network parameters is essential for performing various power system monitoring and control tasks including state estimation, economic dispatch, and contingency analysis. In this paper, we present a novel approach of power network parameter correction wherein we exploit the sparse nature of parameter errors. Parameter error correction using multi-period SCADA or PMU measurements is formulated as a sparse unsupervised regression problem, and an efficient iterative reweighted algorithm is developed to solve the problem. The proposed approach is also generalized to handle parameter errors and topology errors at the same time. The effectiveness of the proposed approach and comparative evaluation is demonstrated using the IEEE 14-bus and 57-bus test cases
Leveraging Structures of the Data in Deep Learning
Get PDFThe performance of deep learning frameworks could be significantly improved through considering the particular underlying structures for each dataset. In this thesis, I summarize our three work about boosting the performance of deep learning models through leveraging structures of the data. In the first work, we theoretically justify that, for convolutional neural networks (CNNs), neighborhoods of a pixel should be redefined as its most correlated spatial locations, in order to achieve a lower generalization error. Based on the correlation pattern, we propose a data-driven approach to design multiple layers of different customized filter shapes by repeatedly solving lasso problems. In the second work, we address the problem of scale-invariance in deep learning. We propose ScaleNet to predict object scales. Through recursively applying ScaleNet and rescaling, pretrained deep networks can identify objects with scales significantly different from the training set. In the last work, we perform an extensive study on PointConv based frameworks to tackle the problems of scale \& rotation invariances in point cloud convolution. PointConv is a novel convolution operation that can be directly applied on point clouds, and achieves parity with 2D CNNs in terms of formulation and performance. It takes coordinates of points as inputs to generate corresponding weights for convolution. We identify two effective strategies -- first, for point clouds converted from regular 2D raster images, we replace the multi-layer perceptrons (MLPs) based weight function with much simpler cubic polynomials, and achieve more robustness and better performance than traditional 2D CNNs on MNIST dataset. Next, for 3D point clouds, we introduce a novel viewpoint-invariant (VI) descriptor utilizing geometric properties between a center point and its local neighbors, as the additional input to the weight function. Integrated with the VI descriptor, we not only significantly improve the robustness of PointConv but also achieve comparable or better performance in comparison to the state-of-the-art point-based approaches on both SemanticKITTI and ScanNet