62588 research outputs found
Sort by
Architectural Modification of Polyamides by Anhydride Addition
No full textCyclic anhydrides are widely used in polyamide blends and composites but only as a functional component of larger macromolecules or inorganic particles. Few studies have been conducted on the effects of cyclic anhydrides in their small molecule form on polyamides. The aim of this research is to gain a greater understanding of the interactions between polyamides and cyclic anhydrides such that higher performing, easier to process, and more cost-effective polyamides, polyamide-based blends, and composites can be created. In this work, the effect of cyclic anhydride addition to polyamides in the melt state was evaluated. The unmodified and modified polyamides were assessed though a combination of rheological, chromatographic, thermal, and mechanical analysis. It was discovered that the addition of molecular cyclic anhydrides on polyamides is quite different than the perceived effect accepted in the current state of art. Mechanisms for the interactions between these anhydrides and the polyamides studied were proposed and ultimately it was postulated that reactive extrusion processes with these raw materials could be employed to reliably adjust a polyamides architecture and molecular weight in a continuous process
Examining the privacy risks in systems of data collection and inference using the concept of informational friction
No full textTraditional models of privacy rely on binary distinctions between public and private spaces, and sensitive or nonsensitive information types. However, these formulations of privacy are not able to fully account for the possibility of inferring information from online data that is not explicitly disclosed. Internet users are under near constant digital surveillance, and companies and other institutions have invested massive resources into collecting and analyzing this data in order to gain information about their users. This dissertation puts forth a new theoretical lens, informational friction, that is well-suited for modern algorithmic privacy threats. Informational friction (IF) is a qualitative measure of how easily information flows from one party to another, whether that information is acquired or inferred.Chapter 2 uses this conceptualization of privacy to highlight risks inherent to the technologies of police facial recognition algorithms and Facebook\u27s targeted advertising model. These systems allow powerful parties to acquire detailed information on citizens by analyzing data that is considered "public": surveillance footage and web activity, respectively. The reduction in (IF) causes a permanent reduction in the amount of privacy people can achieve.Chapter 3 describes an experiment which involved building a machine learning algorithm (stylometric author matching) with the potential to deanonymize members of a popular internet drug harm-reduction forum. By building and testing the system, with cooperation of website staff and members, we are able to characterize the reduction in IF caused by the use of stylometric algorithms on public internet posts.Following the construction and testing of the the results of the previous experiments were visualized and presented to members of the community in a series of qualitative interviews. Chapter 4 shows how some of the beliefs and behaviors of our participants regarding the role of algorithmic systems in privacy are more described as reactions to, and efforts to control the level of IF they experience when posting online. This type of analysis adds further nuance and context to the observation that many internet users appear resigned to or cynical about privacy threats from certain powerful entities. It also reifies certain behaviors as relevant to privacy because they are intended to increase IF, whereas conventional privacy models would overlook these strategies.Finally, the dissertation concludes by aggregating the results of the previous chapters into a set of suggestions for future researchers, designers, and policymakers. The incorporation of IF into analyses of privacy can point towards more informed, robust theories and strategies to protect citizens from the epochal new dangers of digital surveillance and algorithmic data analysis
Probabilistic Life-cycle Assessment and Risk-based Maintenance Optimization of Deteriorating Structures and Bridge Networks
No full textProbabilistic performance assessment of structures and infrastructure systems under deterioration mechanisms is a research area of vital importance. Results of the probabilistic performance assessment serve as the basis for laying out effective life-cycle maintenance strategies for structures and infrastructure systems, which is crucial for the well-being of the society. It should be acknowledged that the importance of probabilistic performance assessment of structures and infrastructure systems has long been recognized by the researchers and decision-makers so that an enormous amount of research work on this topic has been conducted. However, given the scope and complexity of this research area, many research topics within the realm of probabilistic performance assessment are still unvisited or underinvestigated. Under deterioration mechanisms, the optimal life-cycle maintenance strategy is even less investigated than probabilistic performance assessment for structures and infrastructure systems. This dissertation aims to tap into several underinvestigated research topics associated with probabilistic performance assessment, which can ultimately lead to a better understanding of the time-variant performance and how to best allocate maintenance resources for structures and infrastructure systems under deterioration mechanisms. Risk, a performance indicator accounting for both failure probability and failure consequences, is selected as the major performance indicator for structures and infrastructure systems. In this dissertation, risk-based optimization is carried out for both structures and infrastructure systems to determine the associated optimal maintenance strategy under natural hazards.For structures and infrastructure systems, multi-hazards can pose a severe threat to their safety. Interaction between different hazards can add a layer of complexity to the probabilistic performance assessment. Corrosion-enhanced fatigue, which refers to the fact that the presence of corrosion can accelerate the fatigue crack development, is a common hazard facing steel structures. A hybrid approach utilizing both theoretical models and information from structural health monitoring (SHM) activities is proposed to evaluate fatigue details under corrosion-enhanced fatigue. This study also proposes a novel optimal management strategy for ship structures subjected to corrosion and fatigue, as well as for bridge networks subjected to corrosion and seismic hazards. A deterioration mechanism can take various forms. Taking corrosion as an example, different types of corrosion exist for steel structures, such as general corrosion, pitting corrosion, and galvanic corrosion. So far, an overwhelming majority of research works on probabilistic performance assessment of plated steel structures (such as steel bridges and ships, etc.) use general corrosion models, which may not represent the actual corrosion phenomenon for these plated steel structures. In this dissertation, probabilistic performance assessment of steel plates of ship structures is conducted utilizing both general and pitting corrosion models. A major part of this dissertation concentrates on the application of the corrosion-resistant steel in the life-cycle management of steel bridges. Corrosion-resistant steel has a potential to be used in girder replacement actions for the carbon steel bridges that are prone to corrosion attack. To quantitatively assess the applicability of corrosion-resistant steel in the life-cycle maintenance actions for corroded carbon steel bridges, probabilistic performance assessment needs to be conducted on steel bridges consisting of carbon steel girders and/or corrosion-resistant steel girders. When the probabilistic performance assessment for steel bridges is completed, a risk-based optimal maintenance strategy can be determined. This strategy has to determine the minimum expected life-cycle cost for a prescribed target risk level associated with conducting girder replacement actions. The risk-based optimal life-cycle management strategy associated with the application of corrosion-resistant steel can be on an individual bridge level or a bridge network level consisting of multiple steel bridges. In risk-based life-cycle maintenance strategy on a bridge network level, user cost is an important aspect of failure consequences. A novel network user cost estimation approach is proposed to account for the user cost more accurately than the previous network user cost estimation approaches. Finally, research is conducted for a bridge network subjected to both corrosion and seismic hazards to shed light on the risk-based optimal maintenance strategy for infrastructure systems subjected to multi-hazards. It should be noted that risk-based optimal maintenance strategy of bridge structures and bridge networks involves multiple significant factors, such as the service life and correlation of girder resistances. The influence of these factors on the optimal maintenance solutions are explored in this dissertation
Bach-like Tensors on Complete Gradient Shrinking Ricci Solitons and quasi-Einstein Manifolds with Harmonic Weyl Curvature
No full textThis thesis concerns two separate but related topics in Riemannian geometry. The first deals with examining conditions on the Bach tensor, and related tensors that we\u27ll call Bach-like, on complete gradient shrinking Ricci solitons. Huai-Dong Cao and others have obtained results on these manifolds by imposing the condition B=0, but I attempt to generalize some of these results with more elementary tensors. The second part deals with quasi-Einstein manifolds under the condition that the Weyl curvature is harmonic. Results have been obtained from Ricci solitons under this condition, and we have extended these techniques to quasi-Einstein manifolds, obtaining new examples in the process
Understanding Within-School Professional Learning: Teacher Perceptions Of An Action Research Program
No full textEducational action research, a participatory research method, has garnered widespreadattention in scholarship. Previous studies of the intentions and experiences of action researchers have indicated a need for a deeper understanding of within-school action research programs. However, conversations about the utility of educational research have been cyclical. The purpose of this study was to provide an in-depth description and analysis of the implementation of an action research program as a professional learning opportunity at one international school in Brazil. This study employed a qualitative collective case study design to understand teacher perceptions of action research to identify its placement in K-12 professional learning conversations. The case explored was an action research professional learning program. The use of a collective case study allowed for a detailed within-case understanding of varied teacher perceptions of the program. This case study was unique because it studied teacher perceptions in a school with an established action research program. The researcher considered teacher perception of the value of action research, how perceptions changed, how perceptions reflect the value of support, and how teacher context influenced their perception of action research as a professional learning opportunity. The findings of this study provide insight into considerations that schools can take to support action research as a valuable professional learning opportunity. I contend that schools must intentionally align action research with school priorities, plan for necessary pragmatic elements to institutionalize the practice, model and support collective commitment to transform practice, and recognize the impact and outcomes of action research within and outside the school
Biophysical Modeling of the Cytokinetic Ring Anchoring Machinery
No full textIn animal and fungi cells a ring primarily consisting of actin and myosin is formed to assist with cell division during cytokinesis. Formation of the cytokinetic ring is actively studied in fission yeast and many mathematical models have assisted in progressing the field. However, modeling of both ring formation and constriction has been at the micrometer scale and coarser. In fission yeast, anillin related Mid1 and F-BAR protein Cdc15 play key roles in linking the cytokinetic ring to the plasma membrane. Super resolution studies have elucidated the relative organization of ring proteins like Mid1 and Cdc15 in relationship to the plasma membrane. Additionally, large portions of the structure of Mid1 and Cdc15 have been determined crystallographically, while structure prediction methods have continued to improve. Previous studies of Mid1 binding to membrane have indicated both its Pleckstrin Homology (PH) and cryptic C2 domains as potential regions for membrane binding. To investigate Mid1\u27s membrane binding, we applied all atom molecular dynamics methods using the known structure of its C2-PH region and an experimentally determined membrane. We find that Mid1 initially binds through its C2 domain but can further bind through the PH domain. Cdc15\u27s ability to bind the plasma membrane, however, is known to be disrupted by phosphorylation of its intrinsically disorder region (IDR). We performed coarse-grained molecular dynamics simulations of a Cdc15 dimer at different levels of phosphorylation using phosphosites identified by our experimental collaborators in the Gould Lab. We find progressive phosphorylation drives the Cdc15 IDR to localize to the F-BAR ends which may sterically block membrane binding and oligomerization. Our simulations also suggest that dephosphorylated Cdc15 IDR may phase separate, which was verified by the Gould Lab. As the features of Mid1 and Cdc15 investigated are common in other ring proteins, these projects serve more generally as a proof of concept for applying molecular dynamics methods to investigate sub-micrometer scale problems in cytokinesis
Mechanics of Precast Concrete Tunnel Linings and Drop Ceilings under Construction, Service and Fire Loading
No full textThis dissertation aims to evaluate the mechanical performance of segmental tunnel lining systems during construction and service and the fire performance of drop ceilings in roadway tunnels. To achieve this goal, the research focuses on several aspects, including the assessment and design approach development of tunnel lining segments under thrust jack load. This is accomplished via a combination of full-scale experimentation and numerical modeling. Furthermore, the study examines the radial joint rotational behavior of segmental tunnel lining systems through full-scale experimentation. Finally, the research assesses the fire vulnerability of tunnel drop ceilings through thermo-mechanical analysis. Overall, this dissertation seeks to contribute to a better understanding of tunnel infrastructure performance and safety, with potential implications for design and maintenance practices. The first experimental assessment evaluated the performance of precast concrete tunnel lining (PCTL) segments under thrust jack load from the tunnel boring machine (TBM), including assessment of serviceability limit state and ultimate strength, as well as the impacts of supplementary reinforcement and loading eccentricity. Total six full-scale tests were performed with a thrust jack load up to 22.2 MN per pad. At the maximum expected load during installation (5.78 MN per pad), the segments were virtually undamaged, and hairline cracking initiated between the load pads on only one test. At the TBM\u27s ultimate jacking capacity (9.55 MN per pad) surface cracking was observed between and under the load pads; however, the crack width remained below 0.2 mm for all specimens at 9.55 MN per pad. The formation of cracking limit states can be accurately predicted by pre-test linear and nonlinear finite element (FE) models. At overload conditions, the baseline SFRC-only segment exhibited a radial bursting failure at 20.3 MN. The inclusion of supplemental conventional reinforcement does not reduce the level of cracking damage or transverse strain development below the TBM\u27s ultimate jacking capacity; however, at overload conditions, the supplemental reinforcement mitigates cracking and prevents a radial bursting failure. A load eccentricity of 38 mm towards the extrados surface increased the transverse strain and the formation of transverse cracking at a lower load level on the extrados surface. A design approach was then developed to predict the performance of tunnel lining segments under thrust jack load, through two-dimensional (2D) stress diagrams obtained from finite element analysis (FEA), which were then verified by Iyengar\u27s numerical approximation. From the record of tensile strain and cracking propagation obtained from experiments and the tensile stress and strain data retrieved from numerical models, current stress-based design approaches overlook the Poisson\u27s effect caused by stress from other directions and can overestimate the thrust performance of tunnel lining segments. Accordingly, a strain-based design approach for PCTL segments is proposed to accurately and conservatively assess the performance under thrust jack load. This approach is based on the 2D stress distributions along the longitudinal (i.e., loading direction) and transverse directions, and constitutive laws for homogeneous linear isotropic materials. For design purposes, the material properties of compressive strength, tensile strength, elastic modulus and Poisson\u27s ratio should be determined to properly use the strain-based design method. Inverse analysis is also recommended to determine the uniaxial tensile properties of fiber reinforced concrete. The effects of tunnel curvature and loading eccentricity on bursting and spalling tensile stress and strain have also been investigated through FEA. Tunnel curvature has no significant effect on stress and strain caused by thrust jack load. Loading eccentricity causes higher radial bursting stress and strain for less contact between thrust jack pads and raised circumferential joints, and higher transverse bursting strain on the side where the eccentricity is toward. Loading eccentricity has no significant influence on transverse bursting stress, and transverse spalling stress and strain. The second experimental assessment examined the rotational behaviors of full-scale radial joints with 8° skewness at various axial load levels, in order to compare with widely used theoretical solutions, Janssen\u27s model, which assumes the joints to be perpendicular and perfectly contacted as a continuous member with zero tensile capacity. The effects of bolting were also investigated to determine the viability of bolt removal after tunnel construction completes. The rotational stiffness increases with the applied axial load and slowly decrease at larger rotation, while in Janssen\u27s model the rotational stiffness is constant until joint opening, followed by sudden decrease in rotational stiffness. The Janssen\u27s theoretical solutions are closer to experimental results when the axial load higher. At lower axial load level, Janssen\u27s model cannot predict joint rotational behavior properly, for it overestimate the initial rotational stiffness before joint opening and underestimate the rotational stiffness at large rotation. Compared to the available studies that examined the rotational behavior of perpendicular joints, the 8° skewed joint has significant higher flexural capacity and rotational stiffness at large rotation. Bolting has no significant effect on initial rotational stiffness. The fire vulnerability study examines the flexural response of drop ceiling panels in two representative tunnels to standard fire curves and several realistic fires due to vehicular accidents. Standard fire demands per the RWS and ASTM E1529 fire curves are uniformly applied to the ceiling panels, while the heat exposure contours for typical vehicle fires with heat release rates of 30 MW, 100 MW, 200 MW are generated from modeling software CFAST. The FE software SAFIR is used to analyze the thermo-mechanical behavior of the ceiling panels when subjected to various thermal demands from the fire below. The analysis results indicate that drop ceiling panels are highly vulnerable to fire-induced damage and potential collapse both during a fire\u27s active heating phase (from simultaneous loss of capacity and restraint of thermal expansion) and during the subsequent cooling period (due to tension that develops when the permanently deformed panel thermally retracts). The potential for fire-induced damage or collapse of the drop ceiling panels can be mitigated by reducing the fire hazard, removing the drop ceiling, or enhancing the fire resistance of the panels via the application of passive protection or structural hardening