Masters Theses

Permanent URI for this collection

https://hdl.handle.net/10919/9291

Browse

Now showing 1 - 20 of 23330

Impacts of Invasive Asian Jumping Worms on Soil Properties in Turfgrass Systems and their effects on Plant Nutrient Acquisition and Physiological Responses in Greenhouse Experiments
Thompson, Jordan Catherine (Virginia Tech, 2025-08-11)
This thesis examines the ecological impact of invasive Asian jumping worms (Amynthas spp.) in turfgrass systems and controlled greenhouse conditions. In the field, we evaluated jumping worm presence using environmental DNA (eDNA) and assessed associated changes in microarthropod abundance, microbial enzyme activity, soil chemical and physical properties, and turf condition. The study was conducted at four residential sites in southwest Virginia, each with garden bed infestations and adjacent turfgrass lawns. Sampling occurred in May and September 2024. At each site, four fixed transects were established, beginning one meter inside the garden bed and extending ten meters into the lawn. Soil and turf data were collected from one-square-meter quadrats placed at three locations along each transect: within the garden ("Bed"), at the turf edge ("Turf Near"), and six meters further into the lawn ("Turf Far"), resulting in 48 quadrats across all sites. Jumping worms were most frequently detected in turf quadrats adjacent to garden beds, confirming our hypothesis of turf as a zone of invasion. Microarthropod communities, particularly Collembola, exhibited a pronounced early-season increase in jumping worm-invaded areas followed by a late-season collapse. This shift coincided with changes in microbial enzyme activity: chitin-degrading NAG increased after arthropod decline, while nitrogen-releasing LAP decreased most in jumping worm-infested garden soils. Soil structure was primarily affected in garden beds, specifically, large aggregates (>2 mm) increased in jumping worm-infested beds, while mid-sized aggregates (0.25–2 mm) decreased. Soil nutrients remained largely stable, though phosphorus and organic matter showed seasonally dependent shifts linked to jumping worm presence. Firth logistic regression identified organic matter as a significant predictor of jumping worm detection, though this relationship reversed over time, likely reflecting resource exhaustion or outward migration into turf. Turfgrass condition appeared visually unaffected. Collectively, these findings indicate that while aboveground vegetation may appear intact, significant biological restructuring is underway below ground. In a complementary greenhouse experiment, we tested whether invasive jumping worms alter soil conditions in ways that influence bean physiology. We collected jumping worm-free topsoil from a local farm and divided it evenly into two 151 liter bins. Both bins received identical amendments of coconut coir and compost. One bin was inoculated with 80 live jumping worms; the other remained jumping worm-free. These bins were kept moist and left undisturbed for 12 weeks to allow soil conditioning. No fertilizers were added. After this period, we used each soil type to fill ten 3.8-liter nursery pots (n = 20 total), with no jumping worms transferred into the pots. Provider bush beans (Phaseolus vulgaris) were sown directly into each pot, and plants were grown under uniform greenhouse conditions. We repeated this process in a second planting round using the same bins after six months of soil conditioning to test for cumulative jumping worm effects. Across both rounds, we measured developmental stage (BBCH), chlorophyll content (SPAD), stomatal conductance, biomass, tissue nutrients, and soil chemistry. While planting round was the main driver of BBCH and SPAD, jumping worm-conditioned soils in Round 2 consistently showed slightly elevated values, suggesting compounding effects over time. Stomatal conductance declined sharply in jumping worm-free pots between rounds but remained stable in jumping worm-conditioned soils, possibly reflecting moisture or nutrient buffering. Notably, jumping worm-treated plants produced greater fresh biomass but lower nitrogen content, resulting in elevated C:N ratios. This pattern may reflect a mild dilution effect, where increased growth and water content are associated with slightly lower nutrient concentrations in vegetative tissues. However, carbon and nitrogen concentrations in bean pods remained consistent across treatments, suggesting that plants prioritized allocation of limited resources to reproductive tissues despite reduced shoot nitrogen content. Soil chemistry shifts, especially in calcium, magnesium, potassium, and organic matter, were consistent with accelerated nutrient turnover in jumping worm-conditioned soils.
Identifying Healthcare Access and Enhancing Geospatial Analysis with Generative AI
Sherman, Zachary Harold (Virginia Tech, 2025-08-08)
This thesis brings together geospatial modeling and generative artificial intelligence to address healthcare accessibility and automation of spatial analysis. The first study examines disparities in dental care access across six regions of Virginia by comparing driving and public transit modes for both all dental clinics and those accepting Medicaid. Using a modified two-step floating catchment area (2SFCA) method, the research quantifies access based on travel time, supply-demand ratios, and vehicle ownership. Results show that public transit accessibility is significantly lower and more unequal than driving access, particularly for Medicaid recipients, with variation across regions. Spatial error models further reveal demographic factors, such as poverty, race, and vehicle access, influence accessibility patterns. The second study fine-tunes OpenAI's GPT-4o-mini model to convert natural language queries into executable Python code for geospatial analysis. Trained on over 600 geospatial prompt-completion pairs using Virginia health data, the model achieves an 89.7% accuracy rate, improving significantly over the baseline. It integrates spatial reasoning, fuzzy geographic matching, and modular function calls to reduce execution errors and enhance usability. Together, these studies demonstrate how AI and geospatial science can jointly address inequities in healthcare access while making spatial tools more accessible to policymakers, researchers, and the general public.
Factors Associated with Space Use and Reproductive Success of American Kestrels in Northern Virginia, USA
Wolfer, Caylen Eliza (Virginia Tech, 2025-08-06)
Understanding how breeding-season space use varies across individuals and landscapes can provides insight into the resource requirements of species. My thesis evaluated the space use of American kestrels (Falco sparverius), a species of conservation concern, and tested whether their movement behavior was associated with sex-specific parental roles, local resource availability, and central-place foraging theory as a framework for linking movement to reproductive success. From 2021 to 2024, I used GPS loggers to track 69 kestrels (male = 27, female = 42) in northern Virginia and monitored nesting attempts (n = 48) for reproductive outcomes. Mean range size across stages during the breeding season (incubation, nestling, fledgling, post-breeding, and shifted post-breeding) was 0.51 km² (95% CI: 0.42 – 0.61, range 0.01 km²–4.27 km²), and median distance traveled from nest box was on average 325.26 m (SD: ± 495.36, range 5.82–3628.11 m). A post-breeding range shift occurred in 16% of nesting attempts. Model comparisons using AICc indicated space use varied by sex and stage during the parental care period (nestling and fledgling stages), with males exhibiting consistently larger ranges and distances traveled than females. Female space use increased during the fledgling stage, reflecting shared provisioning responsibilities. By post-breeding stage, both sexes had comparable space use, likely reflecting release from parental duties. Higher availability of meadow and hay within 600 m of the nest box were associated with smaller range sizes and shorter distances traveled from the nest box, although meadow availability increased distances traveled during the fledgling stage. In contrast, increased forest availability was associated with larger median distances traveled from nest box and an increased probability of post-breeding range shifts. Median distance traveled from nest box was not associated with number of fledglings or the proportion of nestlings fledged. However, earlier clutch initiation dates and increased use of pasture were positively associated with fledging success. These findings provide partial support for central-place foraging theory, suggesting that kestrels offset energetic costs of increased movement from the nest box through selective habitat use. This is the first study to quantify breeding-season space use for both male and female kestrels using high-resolution GPS tracking. The results highlight how specific cover types—particularly meadow, hay, and forest—influence kestrel space use, and kestrels who foraged more in pasture had greater reproductive success. Land cover within a 600 m radius of kestrel nest boxes had the largest effect on range size and median distance traveled. Additionally, the relatively small range sizes observed in this study suggests that land cover in the piedmont region of northern Virginia may represent high-quality habitat capable of supporting higher kestrel densities. My results provide land managers and policy makers with information useful for decision making regarding habitat management and informing nest box placement for kestrels.
Improving Integrated Pest Management in Sweet Corn System
Currin, Brian Booker (Virginia Tech, 2025-08-05)
Corn earworm (CEW), Helicoverpa zea (Boddie), (Lepidoptera: Noctuidae) is a common insect pest of sweet corn in the United States. The larvae cause chewing damage on corn ears rendering them unmarketable. Traditionally, growers have controlled CEW in sweet corn using multiple applications of insecticides such as pyrethroids during ear development and/or by planting transgenic sweet corn hybrids that produce insecticidal proteins derived from the bacteria Bacillus thuringiensis. However, over the past decade, CEW populations have developed resistance to both the Cry proteins found in many of the Bt sweet corn hybrids and pyrethroid insecticides, which has made management of this pest difficult. In this thesis I investigate strategies that can help improve pest management for this difficult pest in sweet corn. First, I assessed the efficacy of two novel insecticides that could add different modes of action to the insecticide options in sweet corn; these included a product containing a suspension of nucleopolyhedrovirus particles that are specific to CEW and a product that contains insecticidal peptides derived from spider venom. These were tested in conjunction with Bt transgenic corn to enhance their efficacy. Unfortunately, neither of the insecticides provided any effective level of control of CEW. Additionally, I assessed if sweet corn growers could reduce their frequency of insecticide applications by spraying based on a moth capture action threshold compared with standard frequent insecticide applications applied every 2-3 days. I found that a spray program using thresholds as well as more IPM-compatible insecticides provided similar levels of control while reducing the number of sprays needed. In other studies, I assessed the non-target risks of the different insecticides commonly used in sweet corn. I examined the potential for bifenthrin, spinetoram, and chlorantraniliprole to repel bees and other beneficial arthropods after application in sweet corn; repellency could impact their risk of nontarget effects by reducing exposure. I found that there was no significant effect on the number of times honey bees visited the corn tassels or the number of other beneficial arthropods found. Lastly, I examined the residual toxicity of bifenthrin, spinetoram, chlorantraniliprole, and isocycloseram on lady beetles (Coleoptera: Coccinellidae) after 2, 4, and 6 days in the field. I found there was higher mortality in the bifenthrin and isocycloseram treatments across all three application dates than spinetoram, chlorantroniliprole and untreated control.
The Role of Blackberry Grades and Fermentation Methods on Blackberry Kombucha Quality and Aroma Characteristics from Small-Scale Fermentation
Holst, Alisa Shea (Virginia Tech, 2025-08-05)
Most blackberry farms in Virginia are small operations and seek ways to profit from quality-compromised produce that do not meet the cosmetic guidelines of retail buyers and consumers. U.S. No.1 and U.S. No.2 blackberries differ in physical characteristics like size, color, or insect damage but both are still safe to consume. Utilization of the quality-compromised U.S. No.2 blackberries in innovative ways, like in a kombucha product, could help these growers reduce produce waste as well as increase their revenue. However, the production of kombucha, especially on the commercial scale, requires complex, expensive materials and equipment. Therefore, this study aims to develop a simple, small-scale fermentation strategy for blackberry kombucha to determine if and how blackberry grades and fermentation methods impact the final kombucha quality. First, the physicochemical parameters of U.S. No.1 and U.S. No.2 blackberries were analyzed, including pH, titratable acidity (TA), and total soluble solids (TSS), followed by aroma analysis using headspace-solid phase microextraction (HS-SPME) and gas chromatography-mass spectrometry-olfactometry (GC-MS-O). Then, a blackberry kombucha product was made, in biological triplicates, with either U.S. No.1 or U.S. No.2 blackberries using two different fermentation techniques, involving the addition of blackberries either at the start of fermentation (method 1) or after a week of initial fermentation (method 2). Physicochemical parameters, including pH, TA, sucrose, glucose, and fructose contents, and total phenolic content (TPC), were measured at several time points throughout the fermentation. Additionally, aroma-active compounds were extracted using liquid-liquid extraction (LLE) followed by solvent assisted flavor evaporation (SAFE) and characterized by use of GC-MS-O in the final products. The results showed significant differences (p<0.05) in the pH, TA, sugars contents, and TPC between fermented kombucha from method 1 and method 2. The method 2 fermentation technique yielded a more acidic product with fewer phenolic compounds, more residual sugar, and higher concentrations of notable aroma compounds on the last day (day 14) of fermentation. On the other hand, there were few significant differences (p>0.05) in physicochemical parameters or aroma compounds between U.S. No.1 and U.S. No.2 blackberries or between kombucha fermented from either blackberry grade. These findings suggest that compromised quality of blackberry fruit does not necessarily compromise final quality of kombucha, as well as that the method 2 fermentation lends to a more aromatic product. This small-scale kombucha fermentation approach seems to be a feasible way to increase Virginia growers' revenue by repurposing compromised berry fruits for value-added beverages.
Hands On: Building a Maker Space from the Ground Up
Juarez, Danielle Marcela (Virginia Tech, 2025-08-04)
I began this thesis over a year before its ultimate completion, as I initiated an investigative journey into materials-making with six chosen materials: paper, clay, charcoal, graphite, ink, and watercolor, in the spring of 2024. My only goal at the time was to create materials I could use to present my thesis (an as-yet undecided topic) on. I spent the summer researching, harvesting, refining, and making the materials I set out to make. My work that summer emphasized the power of a tactile connection in creativity and Making, the alchemy in creating Something Useful from something existing, and the human drive to put our hands on what we build and make as opposed to engaging passively as a viewer or consumer. My research also highlighted a clear need for "maker" facilities in the Washington, DC Metropolitan area--colloquially known as the DMV--[D]C, [M]aryland, [V]irginia ("May 2023 OEWS Metropolitan and Nonmetropolitan Area Definitions," n.d.). While there are plenty of fantastic ways to enjoy art and creativity, the DMV is comparatively sparse on resources for making art and being creative. In combining those two halves, both the drive to create and make useful materials with our hands, as well as the pressing need for a place and space to do so, I found an area of architecture in which to investigate. What is the ability of architecture to remain "hands on" in today's virtual and digital landscape? Is there still value and possibility in a building constructed from materials made by human hands, from direct resources? In answering these questions, I developed a campus made by artists, artisans, and craftsmen utilizing site materials to create a Maker space that embodies and nurtures creativity and Making. The goal of this thesis is to create a place for Makers to collaborate, communicate, and connect while using tactile creativity as a means of joy and magic. Instead of using heavy machinery to complete towering buildings, the mark on the landscape is relatively minimal and incorporates resources from the land itself to provide materials for the building structure and components. The organizing and financial source of construction is an artisan's co-op, who could not only participate in making the materials from resources on site, but could also benefit from the campus itself once complete. Built with human touch and invested workmanship from artisans, the campus will be a dedicated place of creativity and making to fill a much-needed dearth of artist resources in the DMV.
Control Flow Merging: A Compiler Transformation to Mitigate Branch Misprediction by Branch Elimination
Ramachandra Sharma, Srinivasan (Virginia Tech, 2025-07-31)
Modern superscalar processors have a high theoretical throughput, being capable of executing multiple instructions per clock cycle. In practice, however, this peak is rarely reached because data, structural, and control flow hazards can stall the pipeline and interrupt execution. Branch prediction is widely adopted in modern architectures to mitigate control flow hazards, yet even sophisticated predictors struggle to handle branches driven by random data. Each misprediction flushes speculative instructions and restarts execution on the correct path, making branch mispredictions a major performance bottleneck. Existing compiler solutions, such as outcome pre-computation and simple predication, are effective only for loop-invariant conditions or small side-effect-free branches. This work introduces Control Flow Merging (CFM), a compiler transformation that eliminates branches by converting control dependencies into data dependencies. CFM systematically replaces if and if-then-else regions with semantically equivalent, predicated instruction sequences, even when handling operations with side effects. Our work targets branch elimination at the compiler level by converting control flow dependencies into dataflow. We introduce the Control Flow Merging (CFM) pass for branch removal which replaces if and if-then-else regions with predicated instructions, while maintaining semantics and safety, even in the presence of instructions with side effects and implement it in LLVM 14. We also evaluate its impact on real-world benchmarks. Our experiments show that CFM can reduce branch mispredictions by up to 99 % and improve performance. Across the entire benchmark set, the pass delivers up to 53x speedup in the best case, and a geometric-mean speedup of 1.57×.
From Conflict to Cohesion: Community Voices on Peacebuilding Initiatives Related to the Farmer-Herder Conflict in Northern Nigeria
Ojo, Emmanuel Ohimai (Virginia Tech, 2025-07-29)
Conflict remains a persistent challenge in many parts of Nigeria, particularly in agrarian regions where disputes over land, resources, and identity often escalate into violence. The farmer-herder conflict in Adamawa State has profoundly disrupted social cohesion, local governance, and economic life. While peacebuilding efforts have sought to restore stability, little is known about how these interventions are experienced and interpreted by those most affected. This study explored the lived experiences of community members engaged in Catholic Relief Services' (CRS) peacebuilding initiatives, with a focus on how these efforts have influenced trust, participation, and daily life in post-conflict settings. Guided by interpretive phenomenology, the study drew on the Human Security Framework and CRS's Social Cohesion Conceptual Framework to examine how peacebuilding was lived by farmers, herders, community leaders, SILC members, and local facilitators. In-depth interviews and focus group discussions were conducted across selected communities in Shelleng and Yola South LGAs. Data were analyzed thematically, producing six core themes and multiple sub-themes reflecting the emotional, economic, and relational dimensions of rebuilding peace. Findings revealed that participants did not perceive peacebuilding as a top-down policy or isolated event, but as a fragile and continuous process embedded in everyday interactions. Central to this process were the restoration of intergroup relationships, the revival of livelihoods, inclusion in community structures, and trust in both traditional and formal leaders. However, concerns about equity, sustainability, and the exclusion of certain groups, especially herders, highlighted ongoing challenges. Participants emphasized that peace is experienced not only through infrastructure or training, but through the ability to live without fear, to participate meaningfully, and to feel recognized and valued. This study contributes to the growing literature on grassroots peacebuilding by centering community voices and emotional realities in post-conflict recovery. The findings underscore the importance of holistic, inclusive approaches that address both material needs and the social fabric of communities. Implications point to the need for continued investment in community-led structures, long-term sustainability planning, and more equitable support across livelihood groups. Ultimately, the study demonstrates how peace is cultivated not in policy rooms, but in borehole queues, market paths, group meetings, and the quiet acts of coexisting again.
Comparison of Gait Characteristics of Horses and the Effect of Farriery and Shoes
Gottleib, Katherine Anne (Virginia Tech, 2025-07-29)
Many equine athletes wear metal shoes to provide protection, support, and traction. The two most common horseshoe materials are steel and aluminum. In some equine disciplines, shoes are removed (barefoot) or changed to a lighter shoe material (aluminum) during an equestrian event to achieve perceived advantages such as more aesthetic limb flight. Shoeing has an impact on performance and health of athletic horses. Thus, the impact of shoe characteristics on performance and welfare of sport horses is coming under increased scrutiny. However, objective information regarding the effects of shoeing on gait characteristics of performance and show horses is scant and there is little scientific evidence to guide decisions about shoeing and its impact on competition fairness and animal welfare. Commonly utilized body-mounted inertial sensor systems can be used to measure symmetry of head and pelvic movement, but do not directly evaluate other gait characteristics important to subjective judging. More recently, triaxial hoof-mounted inertial sensor systems have been used to measure stride characteristics in horses. These sensors yield information about hoof orientation in all three axes and allow more detailed analysis compared to other systems. This technology could be a useful tool to measure changes in gaits of horses under various hoof shoeing conditions and yield information to help guide welfare policy. The objective of our study was to determine differences in gait characteristics pertinent to subjective judging of horses at the trot under various hoof shoeing conditions (barefoot, aluminum shoes, and steel shoes). We hypothesized that hoof shoeing conditions would not affect symmetry of head and pelvic movements as detected with body-mounted inertial sensors or mediolateral hoof deviation during swing phase, stride length, and duration of stride phases detected with hoof-mounted IMUs. We also hypothesized that lower foot-shoe weight (barefoot or aluminum shoes) would result in lower early and late swing phase hoof flight height (arc height a and arc height b, respectively) as detected with hoof-mounted IMUs. A prospective crossover design was used. Twelve healthy, adult, client-owned horses without lameness, hoof or limb abnormalities, or positive response to hoof testers were included. The horses included multiple breeds (Quarter Horse, Thoroughbred, Warmblood, and mixed breed) and were used for several different disciplines (western, eventing, and pleasure riding). Horses were evaluated with body- and hoof- mounted inertial sensors at a trot on firm (asphalt) and soft (sand) surfaces. Evaluations were performed in the following order: with existing shoes (baseline), no shoes (barefoot), aluminum shoes and steel shoes. Hoof balance radiographs (lateromedial and dorsopalmar) were obtained at baseline, barefoot, and after each shoe change prior to data collection. All trimming and shoeing were performed by the same Certified Journeyman Farrier. Data were collected using body-mounted inertial sensors (Lameness Locator, Equinosis, Columbia, MO) and hoof-mounted triaxial inertial sensors (Hoofbeats, Werkman Black Tolbert Tolbert, Netherlands). Data collected included Q score (measure of head and pelvic asymmetry), hoof arc height and lateral deviation, stride length, and mid-stance, breakover, swing, and landing stride phase times. Data were compared among shoeing conditions and surfaces (asphalt and soft footing) and analyzed for statistical relevance. Normally distributed data was analyzed using a mixed model ANOVA. Skewed variables were analyzed using the least squares means and Tukey-Kramer to adjust for multiple comparisons. Significance was set as P < 0.05. Significant differences in arc height between aluminum and steel shoes were detected. Aluminum arc height was significantly smaller in both right and left forelimbs (P < 0.0001) in soft (P < 0.0001) and hard (P = 0.0009) footing compared to steel shoes. Barefoot arc height was significantly smaller on soft footing in both right (P = 0.0297) and left (P = 0.0006) forelimbs compared to steel shoes. Hoof arc height was not significantly different between barefoot conditions and aluminum shoes. No significant differences in Q score, lateral deviation, stride length, and mid-stance, breakover, swing, and landing stride phase times were observed. Limitations of the study included the small sample size, and the variation in horse use and breed. Future research should focus on horses of a single breed and those that are actively competing at high level hunters where the practice of changing shoeing conditions for perceived benefit is more relevant. The null hypothesis that there is no correlation between shoeing conditions and hoof arc can be rejected; there are significant differences in the foot arc height for horses shod with aluminum shoes compared to steel shoes, and horses shod with steel shoes in comparison to barefoot conditions. Aluminum shoes did not change foot arc height compared with barefoot conditions and thus do not support the practice of removing aluminum shoes to compete barefoot in certain phases of competition as no significant differences in gait characteristics were observed. The significant findings of this study suggest further investigation is warranted to determine the impact this has on competition fairness and animal welfare.
Detecting presence Of Malicious Hub in MIMI Protocol for Cross-Platform Messaging Interoperability
Sarvaiya, Harditya Ketan (Virginia Tech, 2025-07-29)
The More Instant Messaging Interoperability (MIMI) protocol enables interoperable group messaging across otherwise isolated services such as WhatsApp, Signal, and Telegram. It routes every Messaging Layer Security (MLS) ciphertext through a central hub that timestamps the message and broadcasts it to all group participants. If the hub is compromised, it can silently drop, delay, or reorder messages, undermining order integrity while leaving end-to-end encryption intact. We introduce a lightweight, Merkle-tree-based audit layer that allows clients to detect such misbehavior. Each client stores every received message together with its hub-assigned timestamp in an ordered list. Clients periodically generate a Merkle proof from this list and broadcast it by embedding the proof in an encrypted application message. Because the hub cannot predict which messages carry proofs, it cannot selectively discard them. Upon receiving a proof, other clients verify it and broadcast their own proofs. Any inconsistency is then propagated to the entire room, creating a non-repudiable record of hub misconduct. A Rust prototype built on OpenMLS was evaluated on a 100-node emulated network. With a client sampling rate of 5%, and a hub attack probability of 10%, the scheme detected 95% of message-drop or reordering attacks within the first 40 messages, consumed only 3 kB of additional memory per client, and required less than 1 ms of client-side processing per proof. The audit's memory requirement grows linearly with room size and requires no changes to the hub protocol, providing a practical, low-overhead path to verifiable message-order integrity in large interoperable messaging systems.
From Vacancy to Vitality: Adaptive Reuse Strategies for an Abandoned Church in Washington, D. C.
Molina, Angela M. (Virginia Tech, 2025-07-28)
This thesis stems from a personal interest in adaptive reuse, particularly within the typology of institutional buildings. In the aftermath of the COVID-19 pandemic, churches—once vital community anchors—have become some of the most affected institutions, facing declining attendance and increasing tax burdens. As a result, many have been left vacant, especially in urban areas like downtown Washington, D.C. The aim of this project is to re imagine one such church, no longer functioning as a religious institution, and transform it into a vibrant center for the performing arts. The goal is not to erase the building's historical and architectural identity, but to explore how its spiritual and spatial character can coexist with a new, inclusive program. This transformed space will serve as a gathering point for the community, professionals, and residents—hosting performances, educational programs, and creative events for people of all ages. Through research and design exploration, this thesis questions: Must a church always remain a church? The findings suggest that while the original function of a building may change, its architectural integrity can be preserved and reinterpreted. Adaptive reuse poses a unique challenge to architects—it demands creativity within constraint, respect for the past, and vision for the future. It is not merely a design strategy, but a sustainable, responsible approach to preserving the built environment while responding to contemporary needs.
Autonomous Vehicle Pose Estimation in GNSS-Denied Areas Using Cross-Track Error Measurements
Moomaw, Andrew Jacob (Virginia Tech, 2024-11-22)
Accurate pose estimation is critical for the safe and efficient operation of autonomous vehicles, enabling precise navigation and control in highly dynamic environments. This study presents a novel approach to vehicle pose estimation using cross-track error (CTE) measurements as supplemental inputs for GNSS denied environments. The proposed method leverages CTE, a measure of lateral deviation from a desired trajectory, to estimate the vehicle’s position and orientation in real-time using an extended Kalman filter (EKF) framework. By integrating CTE into an existing pose estimation architecture, the new system is able to provide vehicle state estimates with a greater accuracy than conventional inertial navigation systems (INS). Simulation studies were conducted on benchmark test sets, demonstrating the method’s effectiveness at reducing pose estimation errors by up to 98% in various GNSS denied situations. This research contributes to the advancement of robust and cost-efficient localization strategies, paving the way for safer autonomous vehicle navigation.
Thermal Modeling of Vacuum Assisted Steam Pasteurization for Improved Product Safety of Low Water Activity Foods
Patel, Sahil Falgun (Virginia Tech, 2025-07-23)
Vacuum assisted steam pasteurization is a common method used to kill harmful pathogens on low water activity foods (LWAFs). During this treatment process it is generally believed that the steam will quickly penetrate through the small gaps within a package of product and condense on the surfaces leading to efficient heat transfer. Through experimentation with a full bag (127 x 178 mm) of product it was found that the steam actually penetrated extremely slowly with unpredictable behavior. To improve the effectiveness of the steam condensation process, a system to force flow through the package was incorporated. Single product experimentation was performed in order to develop a conduction model which predicts the product surface temperature using the measured chamber ambient steam temperature. A number of full bag experiments were conducted with 5 products (whole macadamia nuts, macadamia nut pieces, pumpkin seeds, mustard seeds, and Brazil nuts). During these experiments, the local steam temperatures at the top, middle, and bottom of the package were measured. Computational fluid dynamic (CFD) models were developed for each product to model how the steam penetrates and the transient temperature response throughout the package. The model was validated at a total of three temperatures, 60, 70, and 80 °C. Using the CFD model results as the input for the conduction modeled allowed for the surface temperature of the product within the package to be predicted accurately. A dynamic inactivation model was developed to predict the time needed for reduction of the bacteria Enterococcus faecium (E. faecium) on whole macadamia nuts and was used to predict the inactivation of E. faecium on macadamia nut pieces and pumpkin seeds using their respective dynamic temperature profiles predicted from the thermal models.
Complexity Scaling Laws for Neural Models using Combinatorial Optimization
Weissman, Lowell Meyer (Virginia Tech, 2025-07-22)
Recent work on neural scaling laws demonstrates that model performance scales predictably with compute budget, model size, and dataset size. In this work, we develop scaling laws based on problem complexity. We analyze two fundamental complexity measures: solution space size and representation space size. Using the Traveling Salesman Problem (TSP) as a case study, we show that combinatorial optimization promotes smooth cost trends, and therefore meaningful scaling laws can be obtained even in the absence of an interpretable loss. We then show that suboptimality grows predictably for fixed-size models when scaling the number of TSP nodes or spatial dimensions, independent of whether the model was trained with reinforcement learning or supervised fine-tuning on a static dataset. We conclude with an analogy to problem complexity scaling in local search, showing that a much simpler gradient descent of the cost landscape produces similar trends.
Early Growth Response Protein 2 (EGR2) regulation of B cell differentiation in lupus-prone mice
Islam, Mahfuzul (Virginia Tech, 2025-07-22)
Systemic Lupus Erythematosus (SLE) is a chronic autoimmune disease with an unidentified cause and no suitable treatment. It involves abnormal activation of B cells, which produce pathogenic autoantibodies that contribute to inflammation and tissue damage in multiple organs. Therapeutic depletion of B cells in human lupus patients had mixed success due to the complex heterogeneity of B cell subsets, differences in autoantibody production, and the persistence of antibody-secreting plasma cells (PCs). Our laboratory reported that the transcription factor Early Growth Response 2 (EGR2), a key regulator of immune cell function, was markedly upregulated in murine and human lupus lymphocytes. Further, we reported that conditional deletion of Egr2-/- in lymphocytes in lupus-prone B6/lpr mice suppressed pathogenic anti-dsDNA autoantibodies. Notably, Egr2-/- B6/lpr mice exhibited an increased germinal center B (GCB) cell population but a decreased PC subset, suggesting that GCB cells fail to progress into fully differentiated PCs. This finding reveals a novel and unexpected role of EGR2 in the transition from GCB cells to antibody-secreting PCs. These findings provide compelling evidence that EGR2 functions aberrantly in lupus, contributing to disease pathology by facilitating the generation of autoreactive PCs. To further investigate the novel role of EGR2 on B cells. In this project, we conducted extensive experiments using newly developed CD2Cre-Egr2-/- MRL-MpJ- Fas lpr mice, in which Egr2 is selectively deleted in CD2-expressing lymphocytes primarily in T cells and a subset of activated or mature B cells that express CD2. Our findings show that egr2 deletion in MRL/lpr mice leads to a significant reduction in marginal zone B cells and a trend reduction in antibody-secreting plasma cells, while follicular B cells remain largely unaffected. Germinal center B cells were found to be increased, mirroring previous results observed in Egr2-/- B6/lpr mice. However, we observed only a slight decrease in serum anti-dsDNA autoantibodies in Egr2-/- MRL/lpr mice, which was different from what we observed in Egr2-/- B6/lpr mice. Collectively, our data indicate that EGR2 plays a pivotal and previously underappreciated role in regulating late-stage B cell development in lupus. These findings suggest that EGR2 may contribute to lupus pathogenesis by facilitating the differentiation of autoreactive B cells into plasma cells, and it may represent a novel therapeutic target for modulating pathogenic B cell responses in SLE.
Investigating Private Drinking Water Quality and Biomarkers of Associated Health Outcomes in Southwest Virginia
Bokmiller, Richard Grant (Virginia Tech, 2025-07-22)
Approximately 1.8 million Americans living in rural communities do not have access to safely managed household drinking water. Recent research has demonstrated high rates of the US Environmental Protection Agency's (EPA) Safe Drinking Water Act (SDWA) violations in rural areas and heavy dependence on private systems that are not subject to SDWA monitoring and treatment requirements. Few studies concurrently examine both the quality of private drinking water and associated biomarkers of disease. This relative lack of data is particularly acute in Central Appalachia, a region defined by high poverty, aging public infrastructure, and high rates of private water supply. This study aims to measure the prevalence of health outcomes associated with drinking water from a private or environmental source in a rural, Central Appalachian region in Virginia. To accomplish this goal, participating households with drinking water supplied by a private well, spring, or environmental source (e.g. roadside springs) participated in a three-step process: surveys administered by trained teams, drinking water samples collected at the point of use, and health outcome measurements. Water samples were collected from the point of use without flushing, after sanitization and a 5-minute flush, and from any non-bottled, alternative drinking water sources. These samples were analyzed for fecal indicator bacteria, specific enteric pathogens, metals, nutrients, and other inorganic ions. Though private sources are not subject to EPA regulations, water quality results were compared to US EPA maximum contaminant levels (MCLs), secondary maximum contaminant levels (SMCLs), and health reference levels as benchmarks. Measurements of health outcomes included immunoassays of saliva samples and blood pressure measurements. At least one sample from ~73% (n = 11) of eligible households contained detectable total coliforms and ~27% (n = 4) contained detectable E. coli. One quarter of eligible households (n = 3) had water containing specific enteric pathogens, all of which were supplied by a private spring. No samples exceeded health-based standards (MCLs) for inorganic constituents. Only one individual reported experiencing symptoms of enteric waterborne disease, which coincided with detection of Aeromonas in their primary drinking water source. However, the participant also noted a change in medication as a confounding factor for their symptoms. Blood pressure measurements revealed high prevalence (~69%, n = 11) of hypertension despite fewer participants (25%, n = 4) reporting the condition, reinforcing the need for measured health outcomes. Due to recent changes in federal funding and hiring practices, saliva samples were unable to be analyzed and strong epidemiological associations between private drinking water use and enteric disease remain out of reach. Despite this, biological contamination of private and environmental drinking water sources remains prevalent in Central Appalachia.
Constraining the Radial Structure of Seismic Attenuation using Multiple S-Wave Datasets
Irumhe, Emmanuel Paul (Virginia Tech, 2025-07-18)
Seismic attenuation, quantified by the quality factor Q, provides crucial constraints on Earth's thermal and chemical structure due to its strong sensitivity to temperature and presence of volatiles. This study investigates the radial Q structure of Earth's mantle using amplitude measurements from core-diffracted S waves (Sdiff) and core-reflected S waves (ScS), along with direct S and multiply reflected S waves (SSSS).We analyze 7,598 Sdiff, 463 ScS, 5,940 S, and 1,726 SSSS measurements from 448 earthquakes recorded globally between 2009-2017, significantly improving the ray path coverage and the resolution of the seismic quality factor (Q) structure in the lowermost mantle compared to previous studies. Our results reveal that Sdiff amplitudes steadily increase with epicentral distance and exceed those predicted by the Preliminary Reference Earth Model (PREM), indicating structural complexities in the lowermost mantle not captured by existing 1-D Q models. Linear inversion of amplitude measurements unaffected by mantle triplications confirms a high-Q region in the uppermost lower mantle (600-900 km depth), consistent with previous findings from Zhu et al. (2022). However, forward modeling demonstrates that attenuation variations alone cannot explain the observed distance-dependent Sdiff amplitude increase, even when assuming purely elastic conditions (Q = ∞) in the lowermost 500 km. Instead, our modeling indicates that a thin (∼25 km) low-velocity layer with approximately 4% velocity reduction at the base of the mantle best explains the observations. This layer acts as a waveguide, trapping diffracted energy and producing the observed amplitude enhancement. These findings reveal significant structural complexity in the lowermost mantle, suggesting a more heterogeneous core-mantle boundary region than previously recognized. The thin low-velocity layer may reflect enriched iron content and unique physical conditions associated with high temperatures, phase transitions, and chemical interactions with the outer core.
Enhancing Perception Systems using V2V Sensor Fusion
Gwash, Ansh Sundeep (Virginia Tech, 2025-07-17)
With the surge in popularity of autonomous vehicles that depend on complex perception systems to make safety critical judgments, it is necessary to test and improve them. One of the ways this can be done is through vehicle-to-vehicle communication. This concept has been around for decades but was first standardized in 2010. Since then, there have been many hurdles in the path to applying this technology. Security, reliability, latency, and cost are the main reasons for the slow growth in this space. Another main problem is the lack of compelling applications that make overcoming these limitations worthwhile for industry. Autonomous Vehicles rely on a number of sensor types, with the most common ones being Cameras, Radars, and LiDAR. The detections from these three sensors are fused into a track list that can be used to plan and control the vehicles movements. This thesis proposes a system to introduce data from Vehicle-to-Vehicle messages into this fused track list. This extra information can be beneficial in cases when the onboard sensors are occluded or have low visibility. City and highway driving scenarios and Software-in-the-Loop testing is used to evaluate the proposed fused track list.
Utility of Ion Mobility – Mass Spectrometry in Peptidoglycan Structural Determination
Harris, David Orion (Virginia Tech, 2025-06-17)
The peptidoglycan (PG) found within bacterial cell walls is a complex megastructure with components that are difficult to identify without significant time commitment. Potential isomers, modifications, and cross-linking patterns within whole cell digests lead to complex samples that typically require manual analysis. The introduction of a method for streamlining the identification of peptidoglycan components (muropeptides) is needed to develop a better structure identification workflow. This work explores trapped ion mobility – mass spectrometry as a tool to provide muropeptide structural information through analysis of ion mobilities and collision cross section (CCS). PG of two filamentous cyanobacteria, Nostoc commune and Limnospira platensis, were chosen for this work as they have not been characterized before now despite commonplace use as sources of nutrition. The analyses revealed that muropeptides can be organized into distinct regions of similar cross-linking structure by comparing their CCS and m/z values. Both species possess typical Gram-negative structures in-line with their classifications as Gram-negative species. The PG of L. platensis was found to contain agmatine as a modification to the D-iso-glutamic acid residue within the peptide stem. This modification was found to exist within most (~90 %) of its identified muropeptides, indicating that agmatine may play an important role in L. platensis growth and development.
“I Wanna Be the Very Best Like No One Ever Was” in Every Situation: Developing and Validating Achievement Motivation SJTs Across Contexts
Huynh, Christopher (Virginia Tech, 2025-04-21)
Achievement motivation, a key predictor of performance across life domains, has traditionally been studied through fragmented theoretical frameworks and self-report measures that neglect contextual nuance. This thesis proposes a comprehensive, multidimensional model of achievement motivation, operationalized through situational judgment tests (SJTs) that systematically manipulate context. five SJT scales were developed: three measuring goal orientation (mastery-approach, performance-approach, and performance-avoidance) across work, school, and fitness contexts, and two measuring mental toughness (task persistence, emotional control, and utilization of feedback) in school and fitness contexts, with an existing validated SJT used for work. Using exploratory factor analysis (EFA), the study found support for context-sensitive structures of both constructs. Correlational analyses indicated moderate convergence with existing self-report measures and inter-contextual consistency, suggesting that achievement motivation is partially trait-like but also significantly influenced by context. The findings support a chronic preference model and highlight the value of ecologically valid assessments for capturing the dynamic nature of motivation.

Browse

Recent Submissions