UBIR Repository (Univ. at Buffalo)
Not a member yet
5736 research outputs found
Sort by
A Numerically Effective Methodology for Risk-Based Design of Control Systems for Stochastic Structures
No full textPh.D.The full text PDF of this dissertation is embargoed at author's request until 2022-02-19.Failure of structural systems under the action of dynamic loads can be alleviated via active and passive control techniques. The structural systems, control systems, and dynamic loads typically show significant variations in practice. If the uncertainties in structural and control systems and the randomness of dynamic loads are not accounted for properly, the designed controlled systems might fail. To this end, a methodology for solving an Optimization Under Uncertainty (OUU) problem is developed in the present work that will rigorously account for various uncertainties and result in a cost-effective design of systems that satisfy user-prescribed reliability. The methodology is devised using two novel numerically effective schemes that are based on the Importance Sampling (IS) technique. The methodology requires only a single set of high resolution simulation results, which are reused throughout the optimization process. Numerical results indicate that, for a given sample budget, the proposed methodology is likely to ensure a feasible sub-optimal solution of an OUU problem. Using OUU problem formulations, risk-based methodologies for design of active and passive control systems are also developed in the present work. The OUU problem formulation for risk-based design of active control systems is based on the Quadratic Partial Eigenvalue Assignment (QPEA) technique, which is used here within a probabilistic framework for the first time. The methodology for risk-based design of passive control systems is first developed for generic passive dampers and then tailored for nonlinear viscous fluid dampers. Numerical illustrations using a few linear and/or nonlinear civil and aerospace structures, subjected to random dynamic loads, indicate that the control systems designed via the proposed methodologies are likely to ensure satisfactory performance of structural systems. Hence, the proposed risk-based design methodologies might be useful for practical design of various active and passive control systems.**To request an accessible version of the file(s) associated with this item, contact [email protected]. Please include the item's persistent URL [http://hdl.handle.net/. . .] in your request.*
Three Essays on the Impact of Digital Health Interventions: A Multi-Stakeholder Perspective
No full textPh.D.The full text PDF of this dissertation is embargoed at author's request until 2022-02-19.Healthcare costs in the U.S. have been rising while the quality of and access to healthcare have been lacking. Digital health interventions are now viewed as a partial solution to addressing the triple aim of achieving better quality, lower cost, and greater access. The term “digital health” has been narrowly defined as “internet-focused applications and media to improve medical content, commerce, and connectivity.” However, the term’s definition has evolved such that its contemporary version includes a broader range of technologies and interventions with which to improve quality, increase access, and lower costs. At one end of the spectrum, digital health consists of innovative enterprise applications used in the delivery and management of healthcare, including electronic health records (EHRs), the Health Information Exchanges (HIE), lab-management systems, and home-health applications. On the other end of the spectrum, it encompasses health-specific social networks, applications, and decision-making tools that employ healthcare analytics, internet of things (IOT)-enabled devices, artificial intelligence, and other emerging technologies. The World Health Organization (WHO), using more-clinical framing, has classified digital health interventions as those for (1) clients, (2) healthcare providers, (3) health systems or resource managers, and (4) data services. This dissertation, structured as three thesis essays, attempts to provide a deeper understanding of the effects of specific digital health interventions on healthcare costs, quality, and access...**To request an accessible version of the file(s) associated with this item, contact [email protected]. Please include the item's persistent URL [http://hdl.handle.net/. . .] in your request.*
Adapting to Consistent Errors in Non-Native Speech
No full textPh.D.When speaking with someone new, listeners quickly pick up on the regularities in their interlocutor’s linguistic system—e.g., their allophony (Dahan et al., 2008), syntactic attachment style (Kamide, 2012), and word choice (Metzing & Brennan, 2003)—and use that information to help process their speech. However, when such information is incorrect and unreliable, listeners no longer use it to facilitate processing (van Heugten & Christophe, 2013). In fact, before a conversation begins, listeners may decide to make less use of linguistic cues when they expect the signal will be unreliable, as in conversations with L2 speakers (Lev-Ari, 2015). This dissertation examines situations where L2 errors are both consistent and reliable, asking whether listeners can take advantage of the regularities in errors to facilitate comprehension. This dissertation looks at the learning of consistent errors in two domains: Syntactic agreement and word choice. For syntactic agreement errors, two visual world experiments show that listeners can learn errors in subject/verb agreement; after hearing sentences like Where *is the pants, listeners eventually come to expect pants as a possible continuation after hearing Where is...?. Experiment 3, an additional visual world study, examines the learning of consistent errors in grammatical gender agreement. In this experiment, native German listeners were exposed to a non-native speaker who consistently mis-assigned grammatical gender to particular nouns (e.g., saying *der[MASC] Kuh instead of die[FEM] Kuh ‘the cow’); similar to Experiments 1 and 2, listeners showed signs of having learned to associate a noun with an incorrect determiner. For word choice errors, Experiments 4 (visual world) and 5 (mouse-tracking) demonstrate that listeners can learn incorrect labels for specific objects (e.g., learn to associate the label cauliflower with a picture of broccoli). In addition, Experiment 5a shows that listeners continue to expect these incorrect labels even when switching to a native speaker, who should be expected to not produce lexical errors. The final experiments look at whether adapting to incorrect labels produced by a non-native speaker (e.g., calling broccoli cauliflower) differs depending on how biased someone is against foreigners (measured by an Implicit Association Task). One mouse-tracking experiment, conducted in a rural, mostly white community in northern Maine, found that less biased individuals were better able to adapt across the experiment compared to more biased individuals. However, a similar experiment, conducted with a more diverse population of undergrads at the University at Buffalo, was unable to replicate these findings. All in all, this dissertation shows that listeners are able to attend to and learn consistent syntactic and lexical errors in order to facilitate the processing of non-native speech. Listeners will even transfer these learned errors to the processing of native speech. While bias against foreigners may play a role in how listeners adapt to errors in non-native speech, more research is required to understand exactly how and in what circumstances this occurs.**To request an accessible version of the file(s) associated with this item, contact [email protected]. Please include the item's persistent URL [http://hdl.handle.net/. . .] in your request.*
Exploiting the Host-Guest Properties of Self-Assembled Metal-Organic Materials for Gas Separation and Water Purification
No full textPh.D.This project is broadly centered in understanding the host-guest properties of coordination-driven self-assembled metal-organic materials (MOMs) such as discrete molecular metal-organic polyhedra (MOP) and polymeric metal-organic framework (MOF). The self-assembly of these materials has been widely studied due to their interesting properties including well-defined internal cavities, rigid structures, and modular syntheses. These properties enable their use in a variety of applications including light-harvesting, energy storage, catalysis, molecule sequestration, and gas storage. This work is specifically focused on (1) evaluating the capacity of metal-organic materials for pollution management, i.e. water purification, as well as gas storage and (2) designing composite materials wherein an organic polymer is embedded with these self-assembled MOMs. A suite of analytical techniques including spectroscopy, microscopy, diffractometry, calorimetry, and gravimetry are used in order to fully characterize our MOMs and composite materials. Initial work demonstrated the host-guest chemistry of a Cobalt(II) MOF known as a “crystalline sponge” for its ability to encapsulate organic dyes in water. The insolubility and water-instability of most MOFs limit their application in water treatment technologies unlike MOPs which are stable and soluble in a range of solvents making them more processable as additives to mixed-matrix materials (MMMs). We then studied a group of MOPs as fillers for MMMs with a polyvinylidine fluoride (PVDF) polymer phase and contrasted them against MMM containing MOF-5. We demonstrated the formation of thin, flexible, and homogeneous MMMs when a given MOP was soluble in the precursor solutions. With our ability to incorporate MOPs into polymer to make membranes, we started to investigate the self-assembly of group 11 metallacycles for olefin/paraffin separation due to a precedent that silver metal ions can interact with ethylene and light olefins over related paraffins. We observed that the precipitation of Ag(I) and Au(I) hexagonal metallacycles upon assembly in chloroform/methanol mixtures resulted in high solid-state photo-stability. Finally, the anionic molecular cage, FeMOP, was demonstrated to encapsulate per- and polyfluoroalkyl substances (PFAS) in water. The key findings from these host-guest studies can be applied to future metal-organic materials design and modification for improved cage incorporation in mixed-matrix materials with optimized performance in gas separation and water purification.**To request an accessible version of the file(s) associated with this item, contact [email protected]. Please include the item's persistent URL [http://hdl.handle.net/. . .] in your request.*
Quantifying Analyte-Porous Silicon Interactions
No full textPh.D.This dissertation discusses analyte-dependent responses using porous silicon (pSi) as a medium. This research demonstrates how analytes affect intrinsic pSi photoluminescence emission (PLE). Toluene and methanol quench pSi PLE, while nitromethane enhances pSi PLE. Fourier transform infrared (FTIR) spectroscopy is performed to analyze how these analytes affect pSi surface functional groups and oscillator strengths. Single point Raman microscopy is used to determine how these analytes influence the silicon nanocrystal (SiNC) lattice stress. Depth dependent Raman microscopy is utilized to determine how these analytes affect SiNC stress as a function of depth.**To request an accessible version of the file(s) associated with this item, contact [email protected]. Please include the item's persistent URL [http://hdl.handle.net/. . .] in your request.*
Vergil's Allusions to Homer: Quantifying the Reader's Experience of Aeneid 1
No full textPh.D.Vergil's allusions have been the subject of scholarly interest since before the Aeneid was published. Yet the formal features that constitute a typical Vergilian allusion have not been fully described. This dissertation focuses on describing the formal features of a subset of his intertexts, those with the Homeric Iliad and Odyssey, building on the seminal work of Georg Knauer. It further explores the question of how these particular features prompt recall of the Homeric passages. It pursues these questions through a combination of classical philology, computational linguistics, and cognitive theory.In philological terms, this dissertation demonstrates that the modern reader’s reaction to a potential allusion in Vergil can be predicted based on the passage’s formal similarities to Homer. Whenever new similarities to Homer are discovered, they should be compared against the pattern of described in this dissertation. If similarities to Homer fit the pattern described here, the scholar of Vergil can be more confident in a putative allusion. If the similarities fall outside the pattern, we must ask why Vergil would deviate from his usual method of poetic composition. The findings of this dissertation also have consequences for cognitive science. Embodied models of cognition state that sensorimotor systems in the brain are active during language processing; therefore imagistic language and descriptions of physical action should be important to cross-language allusion. By contrast, connectionist models of cognition assert that language processing takes place using a hidden lexicon that is distinct from sensorimotor systems. As a consequence, connectionist models predict that features like word frequency should correlate with allusion. The experiments of this dissertation show that word frequency is a poor predictor of allusion, while both concrete language and the actions of characters are highly effective predictors. Therefore embodied models of cognition are supported over connectionist models in the domain of allusive language use.**To request an accessible version of the file(s) associated with this item, contact [email protected]. Please include the item's persistent URL [http://hdl.handle.net/. . .] in your request.*
Evaluation of Wastewater Disinfectants Against Microorganisms and Extracellular DNA: A Comparison Between Chlorine, Peracetic Acid, and Combined Peracetic Acid and Ultraviolet Light
No full textPh.D.Wastewater disinfection is important to protect human and ecosystem health. Peracetic acid (PAA) is gaining popularity as a wastewater disinfectant, and has the advantage of being less toxic to aquatic species and forming fewer known disinfection byproducts than traditionally used chlorine. Assessment of the relative performance of PAA and chlorine at different exposures and seasonal temperatures is still needed to promote utility adoption, especially in cold environments and for non-traditional applications such as water reuse. While laboratory studies have shown promising results on PAA's disinfection performance, knowledge gaps still remain regarding scale-up and the potential additive effect of combining PAA treatment with ultraviolet (UV) light. This dissertation examines the inactivation performance of PAA, chlorine (NaOCl), UV, and combined PAA plus UV against fecal indicator bacteria (FIB) used as water quality indicators (specifically fecal coliforms, E. coli, and Enterococcus spp.), somatic coliphage (a virus that infects E. coli), and Cryptosporidium parvum (a protozoan pathogen). In addition to microorganisms, this dissertation investigates the effect of disinfectants on extracellular DNA, which is an important area of research as concern increases that antimicrobial resistance genes (ARGs) released into the environment, in wastewater effluent, could contribute to the spread of antimicrobial resistance (AR) worldwide. The first chapter in this dissertation presents laboratory experiments evaluating the relative performance of PAA, NaOCl, and PAA plus UV at different exposures and seasonal temperatures, and against a range of microorganisms in authentic wastewater. Results showed PAA and NaOCl achieving comparable inactivation of E. coli and Enterococcus spp., with Enterococcus spp. exhibiting more susceptibility to increases in contact time than E. coli. Neither PAA nor NaOCl were as effective as UV at inactivating somatic coliphage, and the combined PAA plus UV was even more effective for inactivation of somatic coliphage. Cryptosporidium parvum exhibited the greatest resistance to disinfection, with all treatments achieving minimal inactivation. The second chapter of this dissertation presents a pilot-scale study of PAA performance at a local wastewater treatment plant. This study showed that increasing the applied PAA C·t resulted in an increase in the log reduction of FIB and somatic coliphage in both secondary and tertiary effluents. The relative effectiveness of increasing the applied PAA concentration versus contact time to improve microbial inactivation varied among the microorganisms. Combined PAA plus UV treatment achieved higher inactivation than PAA only for somatic coliphage in secondary effluent, and for fecal coliforms, E. coli and somatic coliphage in tertiary effluent. The third chapter of this dissertation presents laboratory experiments investigating extracellular DNA degradation by PAA, NaOCl, and PAA plus UV. NaOCl was more effective than PAA at degrading extracellular DNA in molecular grade water (MGW). In secondary effluent, both disinfectants were much less effective, and PAA appeared to be more effective than NaOCl, especially at high C·ts. UV effectively degraded extracellular DNA and had a synergistic effect when combined with PAA. Collectively, this dissertation enhances our understanding of disinfection treatments and provides evidence to support the use of PAA as a wastewater disinfectant capable of disinfecting and degrading a variety of wastewater pollutants, especially when applied in combination with UV.**To request an accessible version of the file(s) associated with this item, contact [email protected]. Please include the item's persistent URL [http://hdl.handle.net/. . .] in your request.*
"Jump Off This Building!" The Ethics of the Headphones in Remote Macao
No full textM.A.Rimini Protokoll's Remote Macao (2017) is an immersive performance in which I participated as an audience-performer. There were no prepared actors in this piece. Around thirty paying audience members put on the headphones, wandered around downtown Macao guided by a disembodied voice in the headphones. As one of them, I committed to the voice, physically responded to it, and obeyed almost all the voice’s directions until the end of Remote Macao. At the final location of Remote Macao, on the rooftop of a skyscraper, the voice released the ultimate order, "to jump off this building!" Obviously, nobody jumped. However, this "suicidal" order suggests a kind of disequilibrium between my commitment to the voice and the commitment return. In this essay, I examine this disequilibrium and the ethical problem regarding the final jump. My commitment to the voice derives primarily from my role as an "audience-performer" in the performance: the augmented involvement as an "audience member" and the physical engagement as an "performer." I argue that the headphones in Remote Macao are a technological subject who encompass a power of hailing. Ethically speaking, this subject could be dangerous because it requests potentially hazardous bodily responses. The headphones construct for the audience-performers a conceptual, imaginative space, in which the headphones seduce and compel the audience-performers to obey its command and to push them to the final dilemma in the name of collectivism. Considering the structure of this essay, first, I introduce my embodied experience as an audience-performer in Remote Macao. Second, inspired by Josephine Machon and Claire Bishop, I analyze the unbalance between the commitment and the commitment return. Third, I investigate the headphones as both an object and a subject by referencing Jennifer Parker-Starbuck and Robert Bernstein. Finally, in comparison to the malevolent authority in Stanley Milgram's social experiment, I examine the ethical problem inhabiting the headphones, and how it pushes the audience-performers to the final jumping in Remote Macao.**To request an accessible version of the file(s) associated with this item, contact [email protected]. Please include the item's persistent URL [http://hdl.handle.net/. . .] in your request.*
Transient Studies of Hot-Carrier Transport in 2D Semiconductors
No full textPh.D.The current carrying capacity of semiconductor devices is controlled by numerous distinct processes, taking place at different timescales and whose origins may be either intrinsic or extrinsic. The dominant intrinsic effect arises from optical-phonon emission by "hot" carriers, thus setting an upper limit for the current-carrying capacity of these devices. Typically, however, this ideal operation is not possible due to several additional, extrinsic mechanisms. Dominant among these are "self-heating" of the dielectric layers surrounding the device channel and injection of energetic (‘hot’) carriers (from the channel) into these layers. These problems are exacerbated in scaled devices, forcing researchers to look beyond conventional semiconductors towards materials such as graphene and other 2D semiconductors. In this thesis, we describe the use of a strategy of repetitive transient pulsing to investigate the details of self-heating and charge trapping in BN-encapsulated graphene transistors. When operating these devices under strongly nonequilibrium condition, we observe that, relative to more standard devices fabricated on SiO2 substrates, the encapsulation shows an enhanced immunity to charge-trapping, the influence of which is only apparent under the combined influence of strong gate and drain electric fields. Next, through the application of a single-shot pulsing strategy to MoS2 transistors, we show how it is possible to mitigate the influence of extrinsic factors and to observe instead a saturated drift velocity (~5 – 7 × 106 cms-1) that is significantly higher than that reported in previous works, in which the influence of self-heating and deep oxide traps could not be excluded. Our study therefore reveals important information on the hot-carrier characteristics of devices based on emergent two-dimensional materials.**To request an accessible version of the file(s) associated with this item, contact [email protected]. Please include the item's persistent URL [http://hdl.handle.net/. . .] in your request.*
Taste and Palatability in the Contemporary Memoir
No full textM.A.This thesis examines the role of embodiment in the contemporary memoir by highlighting two works produced within the last 25 years: Autobiography of a Face by Lucy Grealy and Hunger by Roxane Gay. Primarily, it is concerned with marginalized representations of embodiment and their transformative potentiality: firstly, in their ability to incite and illuminate political action, and secondly, in their ability to appeal to both popular and academic readers. Using cultural studies and theories of pain, abjection, and the grotesque, I argue for each writer’s queering of bodily expectations; it is a cultural critique as much as it as an exercise in reparative reading. Finally, this thesis demands a re-evaluation of what writers are accepted into the literary canon