AUETD (Auburn University)
Not a member yet
9771 research outputs found
Sort by
CONSTRUCTING AND EVALUATING LARGE-SCALE INFILTRATION SWALES FOR RETAINING AND INFILTRATING ROADWAY STORMWATER RUNOFF
Get PDFUrbanization, characterized by the increase of impervious surfaces including roads, parking lots, and buildings, presents challenges for stormwater management. The expansion of impervious surfaces disrupts natural infiltration processes, leading to increased volumes and peak flow rates of stormwater runoff. These changes necessitate effective management strategies and practices to mitigate the negative consequences such as flooding, streambank erosion, and pollution of waterways.
In response, post-construction stormwater control measures (SCMs) that integrate Low Impact Development (LID) and Green Infrastructure (GI) principles are gaining traction. LID and GI SCMs utilize various sustainable processes, such as evapotranspiration, infiltration, filtration, and water reuse, to mimic pre-development hydrology and manage the quantity and quality of stormwater runoff. The Alabama Department of Transportation (ALDOT) has embraced an infiltration-based LID and GI SCM called infiltration swales for highway runoff management.
Infiltration swales are linear vegetated channels that employ an engineered soil media and check dams to minimize surface discharge and reduce peak flow rates. These vegetated systems mimic predevelopment hydrology by promoting infiltration of stormwater runoff into their media and the underlying native soils, potentially replenishing the local groundwater table. This infiltration process is facilitated by the engineered soil media, a component designed to optimize water permeability while utilizing natural materials.
While infiltration swales are a common stormwater management practice in Alabama, their performance can vary. To address this variability, ALDOT partnered with Auburn University on a two-phase research project to develop a modified infiltration swale design based on the existing ALDOT standard, and to evaluate both swale’s infiltration performance. The first phase focused on developing a modified swale design based on the existing ALDOT standard. The second phase, which forms the basis of this thesis, involved large-scale field testing to compare the performance of the ALDOT and modified swale designs at the Auburn University Stormwater Research Facility.
This research aims to evaluate and compare the infiltration performance of the standard ALDOT infiltration swale design to the newly developed modified swale design. Large-scale infiltration swales were constructed, side-by-side, to facilitate controlled experimentation and monitoring. The evaluation focused on infiltration rates and drawdown times under various scenarios designed to assess the influence of external factors on infiltration performance. These factors included variations in rainfall frequency, underdrain valve settings (open vs. closed), initial soil moisture conditions (wet vs. drier), and seasonal variations. Moisture content sensors were also installed within the swale media and surrounding soil at different depths and locations to track the movement of infiltrated water. Additionally, settlement of the swales was monitored from construction completion, and surface storage volumes were measured. This comprehensive approach, incorporating various tests and measurements, allowed for a robust comparison of the overall infiltration performance between the ALDOT and modified swale designs.
The evaluation revealed significant performance differences between the ALDOT and modified infiltration swales. The ALDOT swale displayed a notably lower average infiltration rate of 1.6 ft/day (0.49 m/day) and a longer average drawdown time of 12.25 hours than the modified swale. The modified swale exhibited a higher average infiltration rate of 5.2 ft/day (1.6 m/day) and a considerably faster average drawdown time of 5.06 hours. These results indicate a statistically significant difference in performance between the two infiltration swale designs. Further analysis concluded that decreased rainfall frequency increased infiltration rates by 1.6 times more for the ALDOT swale and 2.4 times more for the modified swale compared to increased rainfall frequency. Drier soils increased infiltration rates by 1.5 times more for the ALDOT swale and 2.3 times more for the modified swale compared to wetter soils. Closed valve underdrain tests outperformed open valve tests for both swales potentially due to seasonal variation. Colder months were associated with slower infiltration rates while warmer months were associated with enhanced infiltration rates. Warmer months showed improved infiltration rates by 1.7 times more for the ALDOT swale and 2.7 times more for the modified swale compared to colder months. Moisture sensor data revealed an interesting contrast. While the ALDOT swale held 10% more surface water volume compared to the modified swale, the infiltrated water traversed the modified swale media significantly faster, reaching the bottom interface with the native soil in an average of only 0.13 hours (7.6 minutes). In contrast, the ALDOT swale exhibited a much slower travel time, with infiltrated water taking an average of 1.8 hours to reach the bottom. Further evaluation showed that there was no settlement observed in both infiltration swales.
Beyond the core findings on infiltration performance, this thesis accomplished several key milestones. A comprehensive literature review explored factors influencing infiltration-based SCMs, a study on the different infiltration-based SCMs used across the United States by different Departments of Transportation, and studies on grass swales and their average infiltration rates. The project then progressed to a detailed field-scale construction phase. The construction process included geotechnical and soil investigation, site selection, excavation, media material placement, moisture content installation, and site stabilization for both infiltration swale designs. Another major milestone involved constructing and calibrating the testing apparatus used for the subsequent experimentation and evaluation of the two swales. These combined efforts ultimately served two primary objectives: to inform ALDOT on the modified infiltration swale design created, and to provide robust evidence from large-scale experimentation to support the findings on the modified swale's infiltration performance compared to the existing ALDOT swale design
Human Pose Estimation for Sports and Health: Action Recognition in Sports Analysis and Sitting Posture Correction
No full textHuman pose estimation and action recognition are pivotal tasks in computer vision with extensive applications in sports analysis and health. This dissertation addresses several key research areas to advance these fields. First, it introduces a cost-efficient method to enhance and adapt generative models, enabling broader application and improved performance in diverse contexts. Additionally, the study proposes a comprehensive dataset, VideoBadminton, designed specifically for fine-grained action recognition in badminton. This dataset includes detailed annotations and a wide variety of action categories, setting a new standard for future research in sports analysis.
Furthermore, the research adapts the CLIP model for action recognition on the VideoBad- minton dataset, demonstrating how multimodal models can be effectively utilized to improve the accuracy and robustness of action recognition in sports videos. The dissertation also applies human posture estimation techniques to sitting posture analysis, contributing to lower back pain research and offering potential benefits in clinical settings.
Collectively, these studies leverage deep learning techniques, including convolutional neu- ral networks (CNNs), recurrent neural networks (RNNs), transformers, and multimodal mod- els, to advance the fields of human pose estimation and action recognition. The findings have significant implications for sports analysis, coaching, and health, providing valuable insights into athletic performance, training program effectiveness, and injury prevention. Additionally, the proposed methodologies have broader applications in areas such as healthcare, gaming, and robotics, where accurate human pose estimation and action recognition are critical
High Speed Pipelined SAR ADC with Time Domain sub-Ranging Fine Quantizer
No full textAnalog-to-digital converters (ADCs) stand as indispensable bridges, seamlessly connecting the tangible realities of the physical world with the boundless possibilities of the digitalized virtual world. In recent years, the successive approximation register (SAR) analog to digital converter (ADC) architecture has become the most popular candidate. However, due to the increase in signal processing speed and the advance of CMOS technology in support of the high-speed digital circuits, the performance requirements of the ADCs have been elevated, demanding resolution beyond 10bits and conversion speeds exceeding 1GS/s. In response, researchers have started exploring the integration of various ADC architectures, including pipeline, flash, and delta-sigma, to achieve the optimal balance of high-resolution, high-speed, and low power consumption. As a result, the hybrid pipelined SAR ADCs architectures have been widely adopted in variety of applications. In addition to expanding voltage domain architectures, time to digital converters (TDCs) start gaining attention because of its’ high linearity and low power consumption. Besides, due to its digital circuit implementations, the TDC conversion speed is significantly increased with the advance of the deep submicron CMOS technology.
This dissertation explores the fundamentals of ADC designs and dives into the advanced theory of quantization in both voltage and time domains. Traditional ADC architectures and circuit implementations are thoroughly examined. Additionally, the fundamentals and architectures of time-to-digital converters (TDC) are presented. Various prototypes are included to analyze existing architectures and introduce the proposed hybrid pipelined SAR ADC with a time-domain sub-ranging fine quantizer. In this proposed work, a SAR ADC is implemented as coarse stage, and it is passively connected to three ring-TDCs. These ring-TDCs are operated as time-interleaved fine stages, utilizing self-generated three-phase non-overlap sub-clocks. This hybrid architecture, with the cross-domain implementation, finally achieves the desired high speed and low power consumption performances. Furthermore, the auto-aligned passive residue transfer solution significantly reduced the pipelining stage’s power consumption and guaranteed robust operation under process, voltage, and temperature (PVT) variations.
The ADC prototype has demonstrated itself as an excellent solution for high-speed and low-power applications, with the capability to evolve alongside advancements in sub-micron technologies. The prototype not only proved that the cross-domain quantization (a combination of voltage and time domains) is an excellent candidate for the advanced communication systems, but also highlighted the advantages of implementing time-interleaved sub-stage in high-speed architectures, which relaxes the ADCs input buffer drivability requirements
Evaluation and Estimation of Frequency Excursion in the Bulk Power System with High Penetration of Inverter-Based Resources
No full textAs the penetration level of renewable energy resources increases, the power system analysis of modern power systems becomes more complex due to the dominance of inverter-based resources (IBRs). The high penetration of IBRs leads to different dynamic responses compared to the conventional power system dominated by synchronous generators (SGs). Therefore, the conventional frequency response analysis that targets the SG-based power system needs revision to take account of IBRs' dynamic behavior within the power system. Considering the inertia constants and droop coefficients from the generator and governor data, the inertial power flow (INPF) and governor response power flow (GRPF) are provided in commercial power system software to estimate the frequency response in the phasor domain. However, this functionality is only limited to the conventional power system targeting the dynamics of the synchronous generators, not considering the behaviors of the IBR resources and their interactions with the existing power system components. Therefore, this paper thoroughly investigates the frequency response and dynamic behaviors of the different types of resources with the different ancillary services by running the dynamic simulation using the IBR generic models in the bulk power system. In addition, this work proposes the estimation models of frequency excursion using the investigated system factors that affect the frequency responses under the different penetration levels of the IBR, so the frequency excursion of the bulk power system with high penetration of the IBR can be quickly and accurately estimated and predicted throughout the proposed system frequency model and the machine-learning model. This prediction model can give substantial values to the system operators to prepare power system contingency, provide enhanced system reliability, and optimize the system specification
Futureproof: Computer Enthusiasts and the PC (R)evolution
No full textThis project is a study of the establishment of the ‘PC Master Race’ identity within the larger
personal computer/desktop computer enthusiast and gamer communities. This group (and its
associated ideology) is traceable to the late 2000s and early 2010s and continues to be active to
present. In this paper I will show that the group’s use of technical jargon and fascination with
specific consumer technologies acts as a cultural mediator between larger enthusiast
communities. Additionally, I posit that these ideas have origins in techno-social processes borne
out of commodity computing machinery developed in the 1990s and 2000s. Here, the framework
established by Donald MacKenzie, and later expanded by James S Small, is useful in
understanding ‘technology’ as a contingent social process, not as a linear trajectory. Paths not
taken, or ‘failures,’ are worth evaluation to understand why a given technology developed as it
did. In the case of the ‘PC Master Race’ there is also a strong socioeconomic component at work
– it is not simply a matter of ‘being an enthusiast’ but instead ‘being an enthusiast in the right
way.’ Divining what ‘the right way’ means, and for whom, is the ultimate purpose of this project.
Introduction: An unattributed image macro quips “[t]ired of looking at Bad Screen. Can’t wait to
get home and look at Good Screen” The implicit underlying point being that many in the modern
work force rely upon desktop, deskside, laptop, luggable, or similar pieces of computing
machinery as their daily workplace, while simultaneously anticipating a return to their home
wherein many leisure activities are also provided by computing machinery. Such computing
machines have become a ubiquitous aspect of many people’s lives in the early 21st century. They are used for communication, business, scientific research, and other general tasks that largely
conform under the umbrella term ‘computing.’ Recognition of this is neither novel nor unstudied,
and represents both a significant amount of scholarship on the part of historians of technology, as
well as a realization of mid-20th century technologists’ visions for the future.1 It is fair to say, the
‘Bad Screen’ is well-studied and well understood as a form of modern technology. However,
what about the ‘Good Screen’? Where do leisure activities fit into the broader discussion of the
history of computing machinery, and of computer science? A survey of the totality of uses for
modern computing machinery in leisure would be not only exhaustive to read, but impossible to
accurately compile given the myriad uses that people have found for computing machinery,
especially machinery connected to the Internet, over the last twenty or thirty years. Therefore,
this project is interested in only one possible leisure activity that people engage in: videogaming.
A comprehensive survey of videogaming would also be prohibitively complex, so an additional
series of scope reductions are necessary to further refine our topic to enthusiast videogaming
from ca. 2002-2012 and the ways in which enthusiasts fashion themselves against non-enthusiast, non-expert users. This period is chosen because it is here that we see the price of self-
styled ‘PC Master Race’ - an Internet-based community of computing enthusiasts who describe
their mission as ‘spreading the Second Golden Age of PC Gaming.’2 Even a brief survey of the
‘PC Master Race’ (sometimes stylized/abbreviated as ‘PCMR’ or ‘pcmr’) subreddit will reveal a
community of over 3,000,000 members that is seemingly obsessed with customized high
performance desktop computing machines, as well as one founded on the rejection of
inexpensive commodity machines and ‘gaming consoles.’3 Also striking is participants (often)
waggish use of explicitly religious language when describing their experience – they do not
simply ‘join’ but instead ‘ascend.’ This is also a group that attempts to directly chart its own
history, citing its founding in 2011, and highlighting a number of sources as ‘foundational’ to
understanding ‘what pcmr is about.’4 This work attempts to get at the motivations for the
formation of the PCMR group, analyzing both their claims as a consumer advocacy group and
enthusiast community, as well as the underlying technological story surrounding the gaming-
oriented computing machinery which they use
Generational Impact on Self-Group Distancing Among the LGBTQ+ Community
No full textThe purpose of this study is to analyze how generational membership impacts the presence and frequency of self-group distancing attitudes, behaviors, and cognitive patterns among members of the LGBTQ+ community. After bivariate testing and multivariate regression analysis of data from the “Generations: A Study of Life and Health of LGB People in a Changing Society United States, 2016-2018" (Meyer, 2023) survey, a couple communities within the LGBTQ+ population showed statistical significance in certain aspects of self-group distancing. Specifically, the White Lesbian and White Bisexual populations returned a statistical impact of generational cohort membership on self-group distancing behaviors and a feeling of community connectedness. Additionally, the White Gay community showed statistical significance in generational membership impact on self-group distancing cognitive patterns in the form of internalized homophobia. Across all communities, however, there was no substantive significance that showed a demonstrable effect of generational membership on overall self-group distancing. Considering this evidence, it appears that generational membership is not as important as a person’s social environment and perceived potential for discrimination to occur. Therefore, the best intervention to mitigate self-group distancing from occurring is to facilitate an inclusive environment where the potential for discrimination or prejudice is either low or communally known to be met with a swift response
Classical Techniques in an Unconventional System: Investigation of the Insect Immune System and Gut Microbiome
No full textThe role of the gut microbiome in development and function of the host immune system has long been an area of interest, investigated across taxa for decades. As mammalian systems can sometimes provide a prohibitive challenge in conducting germ-free studies in many laboratories, it has also been of increasing interest to generate germ-free insects to investigate interactions between the gut microbiota and host immune system. Insects such as Drosophila, a system in which methods have been described to generate axenic animals, have a relatively simple gut microbiome, consisting of only a few species. In contrast, the American cockroach (Periplaneta americana) has a gut complexity approaching that observed in mammalian systems, poising this organism as a promising venture for axenic studies. However, the current field of insect immunology lacks standardized methods of hemocyte naming conventions and a full understanding of immunology in hemimetabolous insects. Therefore, the purpose of this dissertation was to determine how we might employ modern immunological techniques to develop an understanding of conventional Periplaneta americana immune composition and function. As an additional project, I also progressed a working protocol for the generation of axenic cockroaches, as well as hemolymph collection of these insects, in our lab. Using flow cytometry, I characterized hemocytes based on cellular morphology (size and complexity), lectin binding capabilities, and production of reactive oxygen species. LEA lectin binding provided a consistent method of identifying and sorting three distinct hemocyte populations, and ROS assays revealed two populations of hemocytes producing ROS to different degrees. Gene expression is a readily accessible metric characterizing immunological function. Therefore, I also aimed to determine the transcriptional profile of both hemocyte and fat body tissues in response to various bacterial stimuli. I found that hemocytes did not have a robust immune response to heat killed S. aureus or E. coli; however, both hemocyte and fat body tissues exhibited an immune response to heat killed and live S. marcescens. My evidence of differential gene expression in these tissues marks the first transcriptional evidence of key components in the IMD pathway in P. americana. Collectively, my dissertation provides fundamental insight into the American cockroach immune system at a cellular and transcriptomic level, providing both novel findings and the necessary baseline for future axenic studies
Evaluating Emotion as a Motivating Operation for Residential Staff Behavior
No full textMotivating operations offer a conceptual framework to understand and evaluate emotions within the context of behavior analysis. Emotions are antecedent variables that can have value- and behavior-altering effects on stimulus events as reinforcers and punishers, indicating they are antecedent variables with motivational properties. The conceptualization of emotions as motivating operations provides a framework for the assessment and treatment of behavior. Anger is an emotion particularly relevant in the use of harsh discipline by staff members for children and adolescents who reside in residential facilities. If behavior in response to the subjective feeling of anger is maladaptive it could result in deleterious outcomes for these youth. The current study utilized a translational approach to assess emotion as a motivating operation for staff members’ responses toward residents. Qualtrics randomized undergraduate students into one of three experimental groups. Two groups experienced an emotion induction procedure for either anger or happiness, and the third group recalled the details of their last 24 hrs (i.e., the control group). Directly following the writing procedures, the participants responded to four vignettes in one of four sequence orders depicting the behavior of a resident (two appropriate and two problem behaviors) from the perspective of a residential staff member. The results indicated the induced emotion did not significantly alter responses to resident behavior. Instead, the findings suggested that repeated resident behaviors and order sequences had the greatest influence on the participants’ responses. Additionally, responding to problem behavior had an association with reduced reinforcement for appropriate behavior
Beliefs and Intentions of U.S. Registered Dietitians/Registered Dietitian Nutritionists Toward Providing Breastfeeding Support to Prenatal/Postpartum Mothers
No full textBackground: Given the nutritional benefits and other advantages of breastfeeding, registered dietitian nutritionists (RDNs) must consider their role in supporting prenatal and postpartum mothers, as it pertains to breastfeeding.
Methods: The online Breastfeeding Support Beliefs and Intentions Questionnaire (BSBIQ) was developed and tested to confirm psychometric properties before using it to investigate the beliefs and intentions of U.S. RDNs toward providing breastfeeding support to prenatal/postpartum mothers. A simple random sample of 5,030 RDNs, obtained from InfoGlobalData’s pool of approximately 12,000 of the nation’s 119,249 RDNs, received a hyperlink to the BSBIQ by email via the Qualtrics platform, as well as 471 RDNs employed by a foodservice company who were emailed directly from an email contact list. A priori power analysis determined that 383 cases were required to represent the population of 119,249 U.S. RDNs (95% confidence level; CI = 5), with 137 cases needed to perform structural equation modeling.
Findings: The data analyses were conducted using IBM SPSS software and SPSS AMOS software. Confirmatory factor analysis (CFA) was performed to measure construct validity and determine the model fit. All 18 items of the TPB constructs displayed highly significant loadings (P < .001 for all) with each item demonstrating its the relationship with its latent constructs. The CFA determined that the model fit the data well. Structural equation modeling (SEM) was performed to assess the direct and indirect relationships of the exogenous and endogenous variables on the outcome variable, Intention. The final SEM demonstrated an improved model fit and indicated that Perceived Behavioral Control (PBC), Subject Norm (SubNorm), Level of Interest in completing continuing professional education related to breastfeeding (LevInt), and NotWhite (as compared to white) were all significantly predicting intention through direct effects (P < .001, P < .001, P < .001, P = .042, respectively). The TPB construct PBC was shown to be the best predictor for RDNs’ intentions toward providing breastfeeding support to prenatal and postpartum women.
Conclusion: Results may inform interventions designed to increase the number of RDNs who provide breastfeeding support to prenatal/postpartum mothers, thus improving breastfeeding rates among the U.S. population
List Coloring in Graphs: Constructions Based on a Refined Scale of Choosability
Get PDFA proper vertex coloring of a graph assigns colors to its vertices so that no two adjacent vertices receive the same color. List coloring is a variation of proper vertex coloring where each vertex is assigned a prescribed list of available colors. In 2020, Xuding Zhu introduced a generalization of list coloring called λ-choosability which makes use of integer partitions to categorize list assignments. λ-partitionability is another framework of list coloring that develops naturally out of λ-choosability. All λ-partitionable graphs are λ-choosable, but the converse is not true. In this dissertation, we characterize all complete k-partite graphs which are only λ-choosable when λ is an integer partition which contains only 1's. We show several constructions of graphs which are λ-choosable but not λ-partitionable. We finally show progress towards constructing a counterexample whose purpose is to highlight a key difference between λ-choosability and λ-partitionability