93887 research outputs found
Sort by
Investigation of Fiber Orientation and Mechanical Properties of Pyrolysis Recycled Carbon-Fiber Reinforced Thermoset Composite
Get PDFWith increasing demand of carbon fiber reinforced fiber thermoset composites, establishing a sustainable cycle for these materials becomes crucial. Pyrolysis is a process of reclaiming carbon fiber from thermoset composites by thermally degrading the polymer at high temperatures allowing the fibers to be extracted. Carbon fiber reclaimed through current pyrolysis processes for thermoset composites typically loses its original shape and orientation, making it difficult to reorganize the fibers. This study investigated the feasibility of maintaining the fiber orientations for continuous fiber reinforced thermoset composite during pyrolysis by stitching the carbon fiber layup to a conformable copper mesh during the manufacturing process. By maintaining the carbon fiber lengths and orientation through pyrolysis, an identical part or similar part can be reproduced and significantly mitigate the fiber reorganizing process. This study used the two-step pyrolysis-oxidation process to reclaim the fibers and vacuum assisted resin transfer molding (VARTM) for sample manufacturing. The changes in fiber orientations were monitored over multiple VARTM-pyrolysis iterations using microscopy for plain weave samples, both with and without the copper mesh. The potential contamination within the plies during the pyrolysis process was thoroughly investigated, and approaches to remove it before the next VARTM process were developed. Additionally, the tensile strength and stiffness of both the control and copper mesh samples were measured at each iteration to assess the decrease in structural performance over multiple iterations
Power and Politics in the Media: The Year in C-SPAN Archives Research, Volume 9
Get PDFPower and Politics in the Media: The Year in C-SPAN Archives Research, Volume 9 features articles from multiple disciplines that use the C-SPAN Video Library to explore recent controversies in American politics. Topics covered include Supreme Court nominations, Supreme Court oral arguments, rhetoric on disasters and COVID-19, and the effect of clothing on the approval of women in power. What unites these topics is the unique use of the video record of C-SPAN to explore the intersections of politics, power, rhetoric, and the media in the contemporary United States. Written in accessible prose, this volume showcases some of the most pressing issues today in a variety of political and communication issues while demonstrating video research methodologies
Development of Fluency, Complexity, and Accuracy in Second Language Oral Proficiency: A Longitudinal Study of Two International Teaching Assistants in the U.S.
No full textThis longitudinal case study investigated the developmental trajectories of two participants enrolled in an ITA oral communication course at Purdue University. The course provides language support to the ITA population who play a crucial role in higher education programs in the U.S. (Gorsuch, 2016). Over four months (i.e., one semester), I collected speech data from participants’ responses to different tasks across 14 weeks and surveyed their weekly practice and goals. The multi-dimensional construct of L2 oral proficiency was assessed by eight large- and fine-grained variables representing various subdimensions of fluency, complexity, and accuracy (CAF, Skehan, 1998), including speed fluency (Articulation Rate & Mean Length of Run), breakdown fluency (Mid-unit Silent Pause Duration & Filled Pause Rate), repair fluency (Repair Rate), syntactic complexity (Mean Length of AS-unit), lexical diversity (vocd-D), and accuracy (Error Rate). Dynamic Systems Theory (DST) methods were employed to visualize trends over time (Lowess trend lines, moving min-max graphs), and correlation analyses (Kendall’s Tau for small sample size and moving correlations) were performed to examine both general and local interactions between CAF dimensions and subdimensions.Results showed noticeable improvement in the L2 oral proficiency of both participants, though the extent of progress varied. Andy, an intermediate learner, improved in breakdown/ repair fluency and accuracy, but not in speed and syntactic/lexical complexity. Zach, an intermediate-toadvanced learner, demonstrated progress in almost all CAF subdimensions except for repair fluency. Several noteworthy findings emerged from the study. First, both participants made noticeable gains in accuracy, which can be attributed to formal instruction. Second, the participants exhibited stabilization or regression in speed instead of progress, which may be a strategic allocation of attention and linguistic resources towards more important aspects such as smoothness (represented by mean length of run, (Lennon, 1990)) and accuracy. Thirdly, gains in complexity were observed only in the more advanced learner, suggesting a potential sequence of acquisition. Improving syntactic and lexical complexity may require a certain level of oral proficiency mastery, and both may be compromised in productions by lower-level students like Andy, whose lexical diversity even decreased over time. Fourthly, observed gains aligned well with semester goals and weekly practices, highlighting the importance of goal setting and practices in language learning. Lastly, more significant correlations between CAF subdimensions were observed in Andy, supporting the limited attentional capacity hypothesis (Skehan, 2014). Additionally, dynamics relationships between CAF were observed for the participants, both in strength and direction. Interestingly, Andy’s shift in the relationship between accuracy and speed fluency from competitive to supportive, indicating development.The study is the first attempt to investigate the development of all CAF dimensions from a DST perspective within the context of an ITA program. As an exploratory endeavor, it offers valuable and detailed insights into the L2 speech learning process of two learners within a particular, local setting. These findings have implications for larger-scale group studies in the field of L2 acquisition, as well as for classroom teaching, assessment, and program evaluation within ITA support programs
Beyond Disagreement-Based Learning for Contextual Bandits
No full textWhile instance-dependent contextual bandits have been previously studied, their analysis has been exclusively limited to pure disagreement-based learning. This approach lacks a nuanced understanding of disagreement and treats it in a binary and absolute manner. In our work, we aim to broaden the analysis of instance-dependent contextual bandits by studying them under the framework of disagreement-based learning in sub-regions. This framework allows for a more comprehensive examination of disagreement by considering its varying degrees across different sub-regions.To lay the foundation for our analysis, we introduce key ideas and measures widely studied in the contextual bandit and disagreement-based active learning literature. We then propose a novel, instance-dependent contextual bandit algorithm for the realizable case in a transductive setting. Leveraging the ability to observe contexts in advance, our algorithm employs a sophisticated Linear Programming subroutine to identify and exploit sub-regions effectively. Next, we provide a series of results tying previously introduced complexity measures and offer some insightful discussion on them. Finally, we enhance the existing regret bounds for contextual bandits by integrating the sub-region disagreement coefficient, thereby showcasing significant improvement in performance against the pure disagreement-based approach.In the concluding section of this thesis, we do a brief recap of the work done and suggest potential future directions for further improving contextual bandit algorithms within the framework of disagreement-based learning in sub-regions. These directions offer opportunities for further research and development, aiming to refine and enhance the effectiveness of contextual bandit algorithms in practical applications
Dynamic Network Modeling from Temporal Motifs and Attributed Node Activity
No full textThe most important networks from different domains—such as Computing, Organization, Economic, Social, Academic, and Biology—are networks that change over time. For example, in an organization there are email and collaboration networks (e.g., different people or teams working on a document). Apart from the connectivity of the networks changing over time, they can contain attributes such as the topic of an email or message, contents of a document, or the interests of a person in an academic citation or a social network. Analyzing these dynamic networks can be critical in decision-making processes. For instance, in an organization, getting insight into how people from different teams collaborate, provides important information that can be used to optimize workflows.Network generative models provide a way to study and analyze networks. For example, benchmarking model performance and generalization in tasks like node classification, can be done by evaluating models on synthetic networks generated with varying structure and attribute correlation. In this work, we begin by presenting our systemic study of the impact that graph structure and attribute auto-correlation on the task of node classification using collective inference. This is the first time such an extensive study has been done. We take advantage of a recently developed method that samples attributed networks—although static—with varying network structure jointly with correlated attributes. We find that the graph connectivity that contributes to the network auto-correlation (i.e., the local relationships of nodes) and density have the highest impact on the performance of collective inference methods.Most of the literature to date has focused on static representations of networks, partially due to the difficulty of finding readily-available datasets of dynamic networks. Dynamic network generative models can bridge this gap by generating synthetic graphs similar to observed real-world networks. Given that motifs have been established as building blocks for the structure of real-world networks, modeling them can help to generate the graph structure seen and capture correlations in node connections and activity. Therefore, we continue with a study of motif evolution in dynamic temporal graphs. Our key insight is that motifs rarely change configurations in fast-changing dynamic networks (e.g. wedges intotriangles, and vice-versa), but rather keep reappearing at different times while keeping the same configuration. This finding motivates the generative process of our proposed models, using temporal motifs as building blocks, that generates dynamic graphs with links that appear and disappear over time.Our first proposed model generates dynamic networks based on motif-activity and the roles that nodes play in a motif. For example, a wedge is sampled based on the likelihood of one node having the role of hub with the two other nodes being the spokes. Our model learns all parameters from observed data, with the goal of producing synthetic graphs with similar graph structure and node behavior. We find that using motifs and node roles helps our model generate the more complex structures and the temporal node behavior seen in real-world dynamic networks.After observing that using motif node-roles helps to capture the changing local structure and behavior of nodes, we extend our work to also consider the attributes generated by nodes’ activities. We propose a second generative model for attributed dynamic networks that (i) captures network structure dynamics through temporal motifs, and (ii) extends the structural roles of nodes in motifs to roles that generate content embeddings. Our new proposed model is the first to generate synthetic dynamic networks and sample content embeddings based on motif node roles. To the best of our knowledge, it is the only attributed dynamic network model that can generate new content embeddings—not observed in the input graph, but still similar to that of the input graph. Our results show that modeling the network attributes with higher-order structures (e.g., motifs) improves the quality of the networks generated.The generative models proposed address the difficulty of finding readily-available datasets of dynamic networks—attributed or not. This work will also allow others to: (i) generate networks that they can share without divulging individual’s private data, (ii) benchmark model performance, and (iii) explore model generalization on a broader range of conditions, among other uses. Finally, the evaluation measures proposed will elucidate models, allowing fellow researchers to push forward in these domains
Empirical Essays on Bias-Motivated Behaviour
No full textThis dissertation is a collection of three papers. Each paper constitutes a chapter. Each chapter empirically examines an aspect of bias-motivated behavior in the United States.The first chapter studies the impact of penalty enhancement statutes by state legislatures on the incidence of hate crimes in the United States. Penalty enhancements may deter crime, however, the passing of such laws may also increase awareness among law enforcement officials and increase arrests. Using administrative data on hate crimes and a difference-in-differences method that leverages state-level variation in the introduction of legislation, this paper does not find a significant effect of the state enactment of penalty enhancement statutes on hate-crime incidence rates.The second chapter examines whether election timing and election outcomes affect the incidence of crimes motivated by hate and intolerance. Using administrative data and a difference-in-differences design that compares election with non-election years, I show that hate crimes increase by an average of 28 percent in the three weeks around a US presidential election. This effect is larger in recent presidential elections and when there is no incumbent candidate. Second, using a similar design and cross-state variation in the timing of gubernatorial elections, I find no evidence that these state-level elections affect hate-crime incidence. Third, using regression-discontinuity designs based on vote counts, I find that the number of hate crimes is not affected by presidential or gubernatorial election outcomes.The third chapter studies the impact of presidential and gubernatorial election timing on the level of toxicity present on social media platforms such as Twitter. Together with Sameer Borwankar, I empirically determine the extent to which the toxicity of Twitter content changes during election times as compared to non-election times. We randomly sample Twitter users and collect all tweets made by this sample around election time. We use a difference-in-differences identification leveraging election and non-election years. We further focus on toxic content that is motivated by political polarization and examine various biasmotivation categories that come up in this content as well as the variation in the intensity of toxicity between national and local election times
Essays on Public Policy
No full textIn this thesis, I investigate how public policy influences economic behavior in both a macroeconomic and microeconomic environment. The first two chapters analyze the impact of local policy and its spillovers. The last investigates how fiscal austerity, when forced upon a country by outside authorities.To be specific, the first essay evaluates the return on state business tax credits to mobile firms. Policy makers aggressively compete for mobile firms with tax credits that incentives firms to locate in state with many states paying more in tax incentives than collecting in corporate tax revenue. An important question in this is estimating returns so that local governments are not over bidding for these mobile firms. I construct a novel data set of Kentucky subsides and exploit a policy that increases the subsidy value in distressed counties. This provides an exogenous shift to subsidy generation and circumvents many biases that are uncontrollable in the firm location decision. I find results that are consistent with past papers on the topic that these subsidies generate a three percent increase in employment at the cost of seven thousand dollars per job.The second essay investigates cross-border shopping for lottery goods. An important question in the area of taxation is estimating the extent to which local tax policy can be circumvented by cross-border shopping. This question is particularly relevant now as states have large differences in legality and tax rates on sin goods such as alcohol, cigarettes and marijuana. I contribute to this discussion by exploiting a natural experiment resulting in retailers in Illinois losing the ability to sell popular lottery tickets for eight days. I use this and a difference in difference estimation technique to estimate how lottery players in Illinois crossed the border to play games in bordering states. I find a large fraction of sales that would have occurred in Illinois crossed over. I further show that players in the interior of the state substituted in local Illinois games at a higher rate than players nearer to the border.The third essay explores how forced austerity impacted the Greek economy post the 2010 European financial crisis. There is an extensive literature surrounding fiscal austerity and how governments choose to change their spending and tax polices to achieve fiscal stability and the ramifications on the country’s economy. When outside authorities determine how a country should engage in austerity, they may have a comparative advantage in budget formation or the creation tax policy. On the other hand, local leaders may have institutional knowledge on how government finances should be ordered to maximize economic stability. I explore this tension by constructing a synthetic Greece that did not undergo forced austerity and find significant decreases in employment and output compared to the synthetic control
The Statistical Foundations of Line Bundle Continuum Dislocation Dynamics
No full textA first-principles theory of plasticity in metals currently does not exist. While many plasticity models make reference to rules based on heuristic arguments regarding dislocations (the fundamental mediators of plastic deformation in crystals), the scientific community still does not have a theory of dislocation dynamics which can recover even basic features of plasticity theory. Discrete dislocation dynamics, though a valuable tool for understanding fundamentals topics in dislocation plasticity, becomes unusable beyond 1.5% strain due to the line length multiplication inherent in deformation. As a result, it is necessary to develop continuum theories of dislocation dynamics which treat dislocation densities rather than individual dislocations. This thesis examines the foundations of one such continuum theory: line bundle continuum dislocation dynamics, which assumes that dislocations are roughly parallel at every point. First, this assumption is given definite meaning and it is shown from discrete dislocation dynamics data that to be appropriate when modelling dislocation densities on fine length scales (resolving densities on lengths less than 100 nm). Second, it is found that an additional driving force, the correlation stress, emerges from coarse-graining the line bundle dynamics. This correction to the dislocation interactions is dependent on tensorial dislocation correlation functions describing the short-range errors in the products of dislocation densities lying on two slip systems. The full set of these dislocation correlation functions are evaluated from discrete density data with the aid of a novel left-and-right handed classification of slip system interactions in FCC crystals. Lastly, a study of the correlation stress in a representative dislocation system suggests that these stresses are roughly one tenth the magnitude of the mean-field dislocation interaction stress. Taken together, this thesis bridges discrete and continuum models of dislocation dynamics and provides a foundation for future work on a first-principles theory of metal plasticity
Investigating Germinating Seeds as Oxygen Scavengers in Hermetic Storage: Implications for Insect Mortality
No full textHermetic storage systems have gained global popularity for their ability to minimize stored product losses by depleting oxygen. However, relying solely on insects to deplete oxygen in hermetic storage, when this process takes longer, can result in (further) damage to stored commodities. This study was conducted to investigate: (i) the potential of four different germinating seeds (soybean, rice, cowpea, and corn) in scavenging oxygen within hermetic storage systems; (ii) the impact of container volume and the number of germinating seeds on oxygen depletion; and (iii) the effects of germinating seeds on insect mortality and grain quality. Among the crops tested, cowpea, during their fourth, fifth, and sixth germination stages (T4, T5, and T6), depleted oxygen below 5% within 12 hours. The fourth stage of cowpea (T4) was identified as a potential oxygen scavenger due to its shorter germination time and ease of handling. Moreover, increasing the number of germinating seeds resulted in a faster initial rate of oxygen depletion in all-sized jars. Doubling both the volume of the jars and the number of germinating seeds had a similar rate of oxygen depletion. Additionally, an equation was derived to predict the required number of germinating seeds based on data from different numbers of seed and container volume combinations. Relative humidity levels increased to approximately 90% when empty jars were used but remained consistent at 40% when the jars were filled with grains. Furthermore, using 10, 20, and 30 germinating cowpea seeds with stored grains and insects, oxygen levels were reduced below 5% at different time intervals. Complete adult mortality of C. maculatuswas achieved within 3-5 days of exposure, depending upon the number of germinating seeds. 20 and 30 seeds achieved complete mortality within 72 hours, while 10 seeds required 120 hours. As the number of germinating seeds increased, egg counts decreased, and moisture content significantly increased in the treatment involving 30 seeds. Furthermore, no adult emerged after 96 and 120 hours of exposure to normoxia for the 30 and 20 seed treatments, respectively. However, in the 10 seeds treatment, a small percentage of adults (0.29%) did emerge even after 120 hours of exposure
The Evaluation of Modular Manufacturing in Controlled Environment Agriculture for Repurposed Urban Spaces
No full textThis thesis aims to evaluate a Modular Manufacturing (MM) technical approach to Controlled Environment Agriculture (CEA) for cultivating plant food crops in a repurposed urban space. The specific approach was to fit a modular hydroponic CEA system into an insulated cooler box with environmental control to act as a micro plant factory. The feasibility of the approach was evaluated and a benchmark comparison between repurposed urban space and controlled lab environments was produced.Possessing accessibility and affordability to desired quantitatively and nutritious food is a pillar for a healthy lifestyle, yet food insecurity is a growing problem worldwide, in industrial as well as industrializing nations. Food insecurity is defined as “lacking the ability to meet nutritional needs at one or multiple times during the year.” [1] Though Developing countries tend to score poorly on the Food Security Index [2], the issue is common in developed countries as well, where countries like the U.S. Possess a household food insecurity rate of above 10% [1]. Especially, subgroups of the urban population and university students in developed countries are represented at a higher rate concerning food insecurity [3], due to food insecurity’s dependence on socioeconomic factors such as purchasing power and local accessibility.Bringing production close to the consumers or to the Point-of-Need (PoN) would be a valuable tool for supplementing traditional food crop production and increasing access to highquality food for groups exposed to food insecurity. This is especially attractive in densely populated areas and college campuses, where real estate is prime. Bringing production to the PoN does however carry certain challenges, such as severe resource restrictions, which are not present in traditional agricultural production in rural areas where there is vast access to land, water, and plenty of sunlight. Pushing the boundaries of CEA research, technology, and application areas will be crucial for the utilization of nontraditional agricultural land, agricultural resource optimization, and food security improvements in difficult-to-farm environments to facilitate delivery to PoN.Salient outcomes:The salient outcomes of this research were that a MM platform was proven to be feasible for CEA cultivation of food crops in a repurposed urban space as well as a controlled location. Specimens cultivated in a repurposed urban space were shown to have a lower growth rate compared to a controlled location, but the important comparison is to the currently nonexistent productivity in such spaces.Intellectual merit:The MM CEA platform was designed, prototyped, and tested using components-of-the-shelf (COTS) as recommended by frugal engineering methodology [4]. This manufacturing platform was engineered for a case study for repurposing unused “garage space” on the college campus at Purdue University. The platform was further used for a set of studies to evaluate the feasibility of the MM platform and the production efficiency of the platform not only in a repurposed urban space but also across harsh environments across winter-spring seasons. Romaine lettuce cultivars were used as a sample plant for winter and spring studies due to their property as a popular consumable, nutritious, and relatively short growth time for better productivity