Doctoral Dissertations

Permanent URI for this collection

https://hdl.handle.net/10919/11041

Browse

Now showing 1 - 20 of 18057

Knowledge Integration in Convergence Research: A Theoretical and Empirical Investigation
Punjabi, Shruti Rajesh (Virginia Tech, 2026-01-02)
Convergence research has gained traction as an approach to address complex societal and environmental challenges by integrating diverse forms of knowledge across disciplines and sectors. We know relatively little about how such integration is defined, structured, and enacted in practice. This dissertation examines how convergence is conceptualized and operationalized in a large, cross-disciplinary and cross-sectoral project, particularly in relation to knowledge integration (KI). It comprises of three interconnected studies focused on answering an overarching research question: How is convergence conceptualized and operationalized in a National Science Foundation Growing Convergence Research (NSF-GCR) project, particularly in relation to knowledge integration (KI)? Study 1 provided the theoretical and conceptual foundations for this research. It focused on four key objectives: (a) understanding the ontological and epistemological perspective on KI, (b) identifying the meta-theoretical and methodological approaches to KI, (c) identifying dimensions of KI in cross-disciplinary collaborations, and (d) developing a conceptual framework of KI to identify the types of knowledge assembled (inputs), how knowledge is exchanged and integrated (processes), and what knowledge integration yields (outputs). To do so, it combines a scoping review methodology with a cited reference search and synthesized three domains of literature: (1) studies of inter- and transdisciplinarity; (2) studies of knowledge co-production in sustainability research; and (3) studies focusing on factors influencing knowledge integration in the Science of Team Science (SciTS) field. The study identifies eight dimensions of knowledge integration: (1) types of knowledge integrated, (2) competencies and education required to practice knowledge integration, (3) organizational structure, (4) types of actor involvement, (5) stages of collaboration, (6) contextual factors, (7) processes and mechanisms of knowledge integration, and (8) types of knowledge integration outcomes. It further organizes these dimensions into a conceptual framework of KI using an Input-Process-Output (IPO) model by O'Rourke et al., (2016). This framework is intended to function as a heuristic to prompt teams to adapt it to specific contexts, projects, and team configurations. It can also be used as a scaffold for designing and evaluating knowledge integration efforts in diverse collaborative settings. The second and third studies used this conceptual and theoretical framework to understand how the members of a National Science Foundation-funded Growing Convergence Research (NSF-GCR) project understand and practice convergence research. Specifically, the second study examined the social architecture of the team Using Social Network Analysis (SNA) and self-reported measures of members' transdisciplinary orientation (Misra et al., 2015), this study uncovered the types of collaborative communities, broker, and integrator roles that emerged in the team and examined how these communities and roles corresponded to team members' network positions and orientations. The study modeled an undirected, weighted collaboration network using twenty team members' levels and frequencies of collaboration with peers and contextualized the network patterns with open-ended responses on team dynamics. It identified three collaborative communities: a leadership core of experienced integrators, mentor-mentee pairs, and domain anchors who provided technical expertise. Broker (major brokers, information carriers, satellite collaborators) and integrator (cross-cluster, hidden, within-cluster, narrow) role classification revealed that boundary spanning depended on the interplay of personal orientation, opportunity, and project context. Influence was distributed beyond formal leadership, boundary spanning was not determined by seniority, and subgroup expertise and between-group reach reinforced each other. Study 2 further outlined practical applications of SNA for the evaluation and design of scientific teams that aim for knowledge integration across disciplinary boundaries. Finally, Study 3 addressed how knowledge is assembled, translated, and routed across science and policy arenas the NSF-GCR project. Using an abductive research methodology, and participant interviews, story maps, and archival research as methods, this study: (a) understood how convergence is conceptualized and operationalized by academic and professional experts in the team; and (b) elaborated an earlier conceptual framework for knowledge integration elaborated in study 1 (Punjabi et al., 2025). I found that participants understood convergence as their ability to co-reason together under evolving constraints through adaptable representations, with progress marked not by consensus or cognitive reframing, but by artifacts that were routable, durable, and auditable across different venues and audiences. I also identified the conditions under which boundary objects become interpretable and actionable in environmental policy contexts. It also determined how boundary orchestration can yield legible and credible decision-making outputs and offered transferable insights and hypotheses to inform the design and evaluation of future convergence efforts tackling complex environmental challenges. Together, this three-manuscript dissertation offers a layered and evolving account of how convergence is conceptualized and operationalized in practice, especially in relation to KI. I examined KI from three vantage points: (a) as a plural, multidimensional construct in the literature; (b) as a patterned social architecture of roles, ties, and orientations in a convergence team; and (c) as boundary orchestration work through which people co-reason under evolving constraints via shared representations. Across these lenses, the dissertation suggests that convergence is not a single endpoint or a stable state of integrated knowledge, but a capacity of the team that develops and evolves over time.
Regulation of Intestinal Development and Function in Early-Life pigs
Talmage, Alyshia Holly (Virginia Tech, 2026-01-02)
The small intestine's plasticity allows for rapid structural and functional adaptions in response to external stimuli. The early postnatal phase is a critical developmental window, during which diet transitions and stress can have lasting effects on overall health and productivity. Attenuated growth performance and feed efficiency may exacerbate production losses and have significant economic implications on the swine industry. Historically, the industry used in-feed antibiotics as growth promoters, but their ban has left the need for nutritional or pharmacological strategies that mitigate intestinal inflammation. Additionally, knowledge gaps remain regarding how Wnt/ β-catenin and Notch signaling pathways influence intestinal remodeling during the early life of pigs. Collectively, these gaps highlight the need for targeted nutritional or molecular interventions during critical developmental phases during the early life of pigs. Topics examined herein focus on development changes of the gastrointestinal tract. The first study evaluated the effects of different concentrations of cannabidiol (CBD) supplementation on intestinal inflammation and function at weaning. Pre-weaned pigs received one of four treatments: low (10 mg/kg CBD, n=8), medium (25 mg/kg, n=8), high (50 mg/kg, n=8) and control (0 mg/kg, n=8) once a day for 5 days starting one day before weaning. The medium dose showed greater lactase (P = 0.003) and maltase (P = 0.042) activities in the jejunum, and increased lactase activity in the ileum (P = 0.020). Jejunal IL-β concentrations were greater in control pigs (P = 0.014) compared with other treatments, while mRNA abundance of intestinal integrity markers varied among treatment groups, with CLDN4 reduced in low pigs (P= 0.045), ZO-1 reduced across all treatments compared with controls (P=0.012), and CCL2 reduced in high pigs (P=0.018). CASP6 was elevated in low pigs (P = 0.012), and TNFα was reduced in medium pigs (P = 0.031). Morphology and goblet cell measurements had no differences. Overall, CBD supplementation had limited effects on intestinal inflammation and function in newly weaned pigs, but the medium dose showed the greatest changes in the jejunum. The second study focused on early life intestinal epithelial development by analyzing the duodenum, jejunum, and ileum of 70 pigs sampled at 0, 7, 20, 22, 25, and 28 days of age (n=10 pigs/day). Frozen tissues were assessed for brush border activity and found shifts in enzymatic activity coinciding with luminal content changes. Formalin- fixed tissues were used for histological analysis, cell phenotyping, and RNA in situ hybridization. Morphological analysis showed an in increased villus height and crypt depth in early life and decreased postweaning. Expression of Sox9 (duodenum, jejunum, ileum P <0.001) and Ki67 (duodenum P = 0.008; ileum P = 0.039) were increased at birth and postweaning reflecting intestinal growth and remodeling. Sucrase Isomaltase (duodenum P = 0.025; jejunum and ileum P <0.001) increased from birth to weaning, coinciding with changes in disaccharidase activity. Somatostatin (duodenum, jejunum, ileum P <0.001) was greatest at birth and at weaning, reflecting adaptions to changes in luminal contents. Goblet cells (duodenum, jejunum, ileum P <0.001) increased until weaning suggesting heightened mucus production in response to compromised barrier integrity. These results were supported by Ussing Chamber assays of fresh jejunal tissue. Reduced transepithelial resistance (jejunum P <0.001) combined with enhanced glucose (jejunum P <0.001) and glutamine (jejunum P <0.001) active transport at birth and postweaning described increased permeability and increased energy demands. Molecular signaling pathways, Wnt/β-catenin and Notch, were evaluated to better understand how stem cell proliferation and cell fate determination are regulated from birth to weaning. Wnt3 (duodenum P = 0.001) and β-catenin (duodenum, jejunum, ileum P <0.001) expression was increased at birth and postweaning, promoting stem cell proliferation to support intestinal remodeling. Notch-1 (duodenum and jejunum P <0.001; ileum P = 0.009) and Hes-1 (duodenum P = 0.007; jejunum P = 0.004; ileum P <0.001) expression increased at birth and postweaning, driving enterocyte differentiation. Together these findings highlight the rapid and coordinated intestinal remodeling pigs undergo during early life and the molecular pathways that drive intestinal development during early life of pigs, and potential targets to develop nutritional interventions that enhance nutrient utilization.
Optimizing Veterinary Drug Residue Monitoring in U.S. Cattle through Trend Analysis and Risk-Based Frameworks
Al Wahaimed, Abdullah Saud (Virginia Tech, 2025-12-23)
Veterinary drug residues in U.S. cattle continue to raise important food safety, regulatory, and public health concerns. This dissertation integrates three complementary investigations to examine residue trends, evaluate human health risks, and propose an improved framework for national monitoring. First, national sampling data from the USDA Food Safety and Inspection Service (FSIS) National Residue Program (NRP) for 2021–2023 were analyzed to identify frequently detected residues across cattle types. Cattle accounted for 92% of all positive samples across species sampled (sheep and goats, pigs, poultry, and fish), totaling 3,107 out of 3,391 detections. Dairy cows had the highest number of detections within cattle (1,264 out of 3,107). Desfuroylceftiofur, penicillin, flunixin, and several sulfonamides were among the most commonly detected violative residues. Second, a multifactorial risk-ranking model was developed to assess the potential public health risks associated with fifteen high-priority veterinary drugs detected in cattle. The model integrated hazard evidence, exposure frequency, and severity of adverse outcomes across eleven risk domains. Sulfonamides consistently ranked as the highest-risk group, while β-lactams and non-steroidal anti-inflammatory drugs (NSAIDs) ranked as intermediate-risk due to frequent detection despite moderate inherent toxicity. Finally, an adaptive decision support framework was created to improve veterinary drug residue monitoring by incorporating drug-specific, species-specific, and exposure-related determinants into a unified evaluation system. The framework provides clear thresholds for prioritizing high-risk drug species combinations and can support more efficient risk-based sampling strategies aligned with the Food and Agriculture Organization (FAO) and the World Health Organization (WHO), as well as the Codex Alimentarius Commission. Collectively, these studies demonstrate the need for continuous monitoring of frequently detected residues, improved prioritization of high-risk analytes, and an adaptive, evidence-driven approach to sampling. The findings offer actionable recommendations to strengthen residue-control programs, enhance regulatory responsiveness, and protect public health.
Porous Carbon Fiber Functionalization and Thermochemical Degradation of Polystyrene-based Plastics for Resource Recovery
Zhang, Yue (Virginia Tech, 2025-12-23)
This dissertation explores the chemical transformation of polyacrylonitrile (PAN)- and polystyrene (PS)-based polymers into sustainable functional materials and chemical products. The research is motivated by two global challenges: the increasing demand for rare earth elements (REEs) for high technology applications and the environmental issues caused by plastic waste accumulation. First, porous carbon fiber (PCF) derived from PAN-based block copolymer was synthesized and functionalized for the extraction of REEs. Second, the thermochemical degradation of PS and PS-based engineering plastic into benzene and alkylbenzene were explored. In the first project, a greener REE leaching method followed by solid-phase extraction using functionalized PCF was developed. To dissolve REE in NdFeB magnets, citrate-assisted leaching was developed and achieved over 90% leaching efficiency, reducing the use of mineral acid. A block copolymer-based PCF with high surface area and well-controlled porous structure was synthesized and functionalized with diglycolamide (DGA) ligands. Two different functionalization strategies, using silane-modified and azide-modified DGA, were employed. Although functionalization of PCF with both strategies was successful, limited improvement in extraction capacity was achieved. The poor performance was attributed to low ligand loading, partial pore blocking, and strong citrate-REE complex interaction. This study provided PCF functionalization strategies but showed challenges of REE extraction with functionalized PCF as an adsorbent. The second project focused on the degradation of PS into benzene and its subsequent upcycling into alkylbenzene. PS was thermochemically degraded using AlCl3 as a catalyst to produce benzene. The effect of Lewis acid type, solvent, and AlCl3 concentration was investigated. The optimized reaction conditions for benzene production were identified as 100 mol% AlCl3 in cyclohexane at 100 °C for 5 h. The recovered benzene was further upcycled to alkylbenzene by Friedel-Crafts alkylation with 1-dodecene or 1-chlorodedecane. Both one-step and two-step alkylation reactions for alkylbenzene production were explored, but a low alkylbenzene yield was achieved. GC-MS analysis revealed the formation of alkylbenzene isomers and alkane side products, indicating low selectivity and competing side reactions as the limitation of alkylbenzene production. The third project extended the thermochemical degradation strategy to an engineering plastic, acrylonitrile-butadiene-styrene (ABS), containing a PS block. Compared with PS, ABS is a multiphase polymer, containing two other polymer blocks, PS and polybutadiene (PBD), which make the degradation pathway more complicated. Under similar conditions of PS degradation using AlCl3, the benzene yield was lower in ABS degradation. Size exclusive chromatography (SEC) analysis revealed the incomplete degradation with crosslinked structures. The effect of solvent on benzene yield was investigated, showing that the benzene yield was affected by both aromaticity of the solvent and the solvent-AlCl3 interactions. Deuterated benzene as an aromatic solvent enhanced benzene yield due to charge-transfer interaction with AlCl3. Degradation studies on PAN and PBD homopolymers showed that both PAN and PBD underwent partial backbone degradation, forming short-chain fragments. Nitrile groups in PAN remained intact, and PBD underwent crosslinking. These results showed that non-PS blocks altered the degradation pathways and hindered the benzene production. Overall, this dissertation advances the understanding of polymer-derived porous carbon materials and thermochemical plastic degradation for sustainable material recovery. While challenges remain in the improvement of REE extraction capacity and achieving high alkylbenzene and benzene yield, the results highlighted the potential of PCF-based adsorbents for REE extraction and thermochemical degradation pathways for plastic recycling. These studies not only provide fundamental insights into structure-property relationships but also suggest future directions toward greener and more efficient strategies for addressing critical metal recovery and plastic waste management.
Simulating Earthquake-triggered Runout using Higher-order Hydromechanical MPM and PM4Sand
Alsardi, Abdelrahman Munther Kh (Virginia Tech, 2025-12-23)
Earthquake-triggered landslides pose a significant risk to society and infrastructure. Despite the catastrophic consequences, significant questions remain regarding their manifestation and have rarely been analyzed in large-strain numerical frameworks. There is a need to simulate earthquake-triggered landslides within a unified framework that can incorporate geometrical and material nonlinearities, while having the capability to simulate the runout process, from triggering to post-failure cyclic mobility. The Material Point Method (MPM) is a particle-based computational technique that has achieved significant success in simulating geotechnical problems; however, it suffers from numerical errors that have prevented it from realizing its full potential in geotechnical earthquake engineering. The aim of this research is to develop and validate an MPM framework that facilitates a deeper understanding of earthquake-triggered landslides. To achieve this goal, a higher-order multiphase isogeometric MPM framework is proposed. Non-zero kinematic and periodic boundary conditions are developed to simulate shaking table loading and free-field modes. Finally, the PM4Sand constitutive model is implemented in the hydromechanical MPM to enable the model to capture the cyclic behavior of sand, including liquefaction triggering and cyclic mobility. Verification and validation of the framework are performed by comparing the results with those obtained using the Finite Element Method (FEM), analytical solutions including Newmark-type techniques, and shaking table tests. A centrifuge test of a clay embankment overlying saturated sand is simulated using the computational framework. The accelerometer and pore water pressure trends are satisfactorily matched with experimental data, and the deformation mechanism is well captured. The MPM framework proposed in this dissertation presents promising advancements toward a stable hydromechanical large-strain methodology that can capture material and geometric nonlinearity, simulating the entire runout process from triggering to post-failure stabilization to ultimately better understand earthquake-triggered landslide manifestations.
Multi-Fidelity Analysis and Optimization of Lightweight Structures Subject to Significant Fluid-Structure Interaction Effects
Narkhede, Aditya Avinash (Virginia Tech, 2025-12-23)
The motion and deformation of structures such as aircraft, turbomachinery, offshore marine systems, and next-generation prosthetic devices are inherently coupled with the surrounding fluid flow. These interactions, commonly referred to as fluid-structure interaction (FSI), require the simultaneous consideration of the fluid and the structural sub-problems while capturing the effects of coupling at their interface. Motivated by this, the first part of this thesis focuses on developing a coupled simulation framework for modeling the dynamic behavior of lightweight, compliant containment structures subjected to internal detonation-driven fluid loading. These one-time use structures incorporate sacrificial components that undergo large, permanent deformations, while the internal fluid flow exhibits complex shock waves, reflections, and interactions. A three-stage solution strategy is proposed to capture the distinct physics and time scales of explosive burning, shock propagation, and the structural response to the FSI loading. The resulting framework couples compressible computational fluid dynamics (CFD) with nonlinear computational structural dynamics (CSD), and is used to analyze a model problem involving a thin monolithic steel container. The case study shows that accounting for the FSI effects in the analysis of such lightweight containers is essential, as traditional surface pressure approximation methods underpredict the structural deformations, while decoupled simulations overpredict them. Motivated by the excellent energy absorption capacity and low density of cellular materials, the second part of the thesis investigates their application to explosion containment structures to enhance their blast survivability while reducing their structural weight. Specifically, this thesis focuses on metallic foams, which feature an extended plastic plateau during which their cellular structure collapses due to wall buckling or yielding, followed by a nonlinear densification stage where the material rapidly gains strength. To capture this complex behavior, a custom constitutive model is implemented using an extended von Mises yield criterion combined with a nonlinear hardening law. This model is integrated within the CSD solver and used to analyze a model sandwich composite containment vessel featuring foam cores enclosed between thin steel facesheets. Comparative analyses show that the sandwich composite container performs $9$ times more plastic work and dissipates $8$ times more energy from the applied blast load compared to an equal-weight monolithic steel container. These findings demonstrate the potential of metallic foams to improve blast mitigation performance and highlight the need for an optimization framework capable of systematically exploring the structural design variations while accommodating the coupled FSI effects. The optimization of structures in the presence of FSI effects introduces several challenges. The optimization variables, objective functions, and constraints are often relevant to the structure, while the computational cost is dominated by repeated high-fidelity CFD analyses. Moreover, gradient-based optimization is difficult to apply due to the challenges of differentiating through coupled CFD-CSD solvers. These challenges are addressed in this dissertation by first leveraging gradient-free methods, such as genetic algorithms, to avoid complex gradient evaluations. Specifically, the dissertation introduces the SOFICS (Structural Optimization through Fluid-structure Coupled Simulations) framework, which is a modular optimization workflow that integrates our open-source CFD and CSD solvers with Sandia National Lab's optimization toolkit DAKOTA for its implementation of gradient-free optimization methods. The framework supports parallel batch evaluations on large high-performance computing (HPC) machines through DAKOTA's tiling approach and SOFICS's careful load balancing and resource assignments, thereby speeding up the optimization process. The final part of this dissertation presents a novel adaptive multi-fidelity optimization framework that exploits the imbalance in optimization relevance and computational costs in FSI-based optimization studies. The proposed method combines high-fidelity fluid-structure coupled simulations with a lightweight, on-the-fly surrogate model for fluid-induced loads. To maintain optimization relevance, the approach retains the full CSD model throughout the optimization. As the optimization progresses, the CFD data from the early iterations is used to incrementally build and refine the non-intrusive surrogate model based on a nearest-neighbor search and local interpolation technique. A Gaussian-process decision model is also refined progressively to estimate the surrogate error and automatically determine when a coupled simulation is required. It is expected that, as design evaluations cluster near the optimal solution, the accuracy of the surrogate model will naturally improve, leading to fewer CFD evaluations. The effectiveness of this framework is demonstrated through a numerical example involving the shape optimization of a cantilever panel subjected to shock loading, showing substantial reductions in computational cost while maintaining accuracy comparable to optimizations performed entirely with high-fidelity FSI simulations.
Paraspeckle protein NONO regulates active chromatin by allosterically stimulating NSD1
Hsu, Chen-I (Virginia Tech, 2025-12-23)
Epigenetic events refer to heritable changes in phenotypes that occur without alterations in the underlying DNA sequence. Among the major layers of epigenetic control, methylation of histone H3 at lysine 36 (H3K36) and lysine 27 (H3K27) defines euchromatin and facultative heterochromatin, respectively, thereby distinguishing transcriptionally active from repressive chromatin domains. These modifications are catalyzed by distinct families of histone methyltransferases: the Nuclear Receptor–Binding SET Domain Protein (NSD) family for H3K36me2 and the Polycomb Repressive Complex 2 (PRC2) for H3K27me3. NSD1, in particular, plays a crucial role in maintaining euchromatin integrity, and its loss or aberrant activation has been implicated in human congenital disorders, such as Weaver and Sotos syndromes, as well as broad types of cancers, including Diffuse Midline Glioma (DMG) and a subset of Acute Myeloid Leukemia (AML). This dissertation presents three studies focusing on the biochemical properties of NSD1, a large (~295 kDa) H3K36me2 methyltransferase essential for euchromatin regulation. Due to its size and extensive intrinsically disordered regions (IDRs), NSD1 has been difficult to purify and characterize. In the first study, I developed a baculovirus-insect cell system for expressing full-length (FL) NSD1 and SETD2. By isolating monoclonal baculovirus clones and optimizing a single-step FLAG purification protocol, I obtained highly pure recombinant proteins suitable for enzymatic assays. Importantly, both enzymes retained catalytic activities, representing the first successful reconstitution of full-length NSD1 and SETD2 in a defined biochemical system and enabling downstream structural and biochemical studies for the research community. In my second peer-reviewed publication, I investigated the molecular mechanisms that activate and regulate NSD1. Our study revealed that NSD1 requires allosteric activation through the aromatic pocket of its PWWP2 domain, which interacts directly with the nuclear paraspeckle protein NONO. This protein–protein interaction enhances the catalytic activity of NSD1 toward H3K36me2 deposition. Mouse embryonic stem cells harboring mutations within the PWWP2 aromatic pocket exhibit impaired differentiation into neural progenitor cells, a phenotype partially reproduced by NONO depletion. Intriguingly, NSD1 and NONO mutation are found to drive a rare and understudied macrocephaly phenotype in Sotos and MRXS34 syndromes, respectively. Together, these findings uncover a previously unrecognized mechanism of how nuclear paraspeckes regulate active chromatin, provide an insight into the molecular pathogenesis of macrocephaly, and highlight NSD1-PWWP2 as a vulnerability for therapeutic targeting of NSD1-dependent cancers. In my third peer-reviewed review article, I propose a model in which paraspeckles, NONO, and NSD1 cooperate to regulate euchromatin. We speculate on disease mechanisms driven by disruption of this axis and highlight future directions for targeting NSD1 in epigenetic therapy. Our findings reveal an unexpected layer of NSD1 regulation via paraspeckle-mediated allosteric control, with implications for chromatin state transitions during development and disease. In summary, this dissertation establishes a method for purifying full-length NSD1 and SETD2, overcoming longstanding technical challenges. It also identifies NONO as an allosteric activator of NSD1 and proposes a regulatory model linking paraspeckles to euchromatin dynamics. These findings advance our understanding of chromatin biology and provide a foundation for future therapeutic interventions for human pathological conditions.
High-Resolution Time-Synchronized Monitoring and Anomaly Detection in Modern Distribution System
Zhou, Yijie (Virginia Tech, 2025-12-23)
The increasing penetration of renewable generation introduces new challenges to distribution system monitoring, including faster system dynamics and increased harmonic distortion. This dissertation aims to enhance traditional monitoring frameworks by developing high-resolution, cross-synchronized fundamental and harmonic measurement techniques that assist system operators in more effectively extracting additional dynamic information in modern power distribution systems. The first half of the dissertation focuses on developing advanced time-synchronized measurement techniques for identifying system-wide disturbances and dynamic behaviors. In Chapter 2, a cross-synchronized, frequency-adaptive synchrophasor estimation algorithm is developed to generate time-synchronized fundamental and harmonic phasors with high resolution, enabling real-time spectral analysis. The algorithm improves upon the traditional Discrete Fourier Transform (DFT) by employing adaptive windowing to suppress spectral leakage. In addition, its harmonic estimation accuracy is enhanced through the implementation of M-point average filters that mitigate the leakage from the fundamental component. In Chapter 3, an adaptive linear state estimator for unbalanced distribution systems is developed. A novel optimal PMU placement (OPP) scheme is proposed to guarantee complete observability of the system. The linear state estimator further improves its robustness under contingencies by adaptively reformulating its model to account for topology changes. The second half of the dissertation explores applications of the techniques developed in the first half. In Chapter 4, an SVM-based detector for transformer saturation caused by geomagnetically induced currents (GICs) is developed using synchronized harmonic real power derived from the harmonic synchrophasors estimated by the algorithm in Chapter 2. This approach offers a more cost-efficient alternative to direct GIC measurements. Finally, in Chapter 5, a detection and localization scheme for incipient faults is developed based on singular value decomposition (SVD). Faults are detected by tracking abrupt changes in singular values, and their locations are determined by analyzing correlations among the participations of each monitored bus in fault-related singular value variations.
Risk-based Renewal Prioritization Models (RPM) for Potable Water Pipeline Infrastructure Systems
Vishwakarma, Anmol (Virginia Tech, 2025-12-22)
Water pipelines are critical infrastructure assets buried across the United States, responsible for delivering safe drinking water at adequate pressures from source to customers. A majority of these pipelines were installed in the mid-twentieth century without adequate financial planning for future renewal, creating a growing renewal backlog under tight budget and operational constraints. Decades of utility data and practice-based knowledge, combined with advances in Artificial Intelligence (AI) and computational resources, now make it possible to revisit how renewal decisions are made. A review of current water pipeline renewal methods reveals major gaps, including weak integration of risk with decision criteria, ad hoc selection of modeling algorithms without strategic foresight, and limited, often internal-only, real-world validation. This dissertation addresses these gaps by developing and testing an AI-enabled framework for risk-based renewal prioritization of water pipelines. The work has four main goals: (1) developing an AI model to predict the performance and Likelihood of Failure (LOF) of any water pipeline segment on a 0–5 scale, (2) creating an AI model to predict the Consequence of Failure (COF) of any segment on a 0–5 scale, spanning economic, en-vironmental, and social/service impacts, (3) building a multi-criteria optimization model to generate prioritized renewal portfolios that incorporate risk, cost, equity, and delivery con-straints within budget limits, and (4) establishing experimental protocols to evaluate, veri-fy, and validate model results against field inspections, retrospective failures, and expert judgement across multiple utilities. Applied to several U.S. utilities, the integrated LOF, COF, and portfolio models outperform age-based and heuristic baselines on predictive accuracy, calibration, and risk-reduction-per-dollar, while producing more spatially coherent and operationally feasible renewal programs in retrospective tests. Finally, this research evaluates whether the additional effort required for data collection, model interpretation, and governance is justified relative to current utility practices, with tradeoffs assessed in terms of reduced emergency failures and costs, enhanced transparency and accountability in decision-making, and improved public trust. In the short term, the proposed framework supports more cost-effective and defensible capital improvement planning; in the long term, it provides a template for shifting water utilities from reactive, break-driven repairs to proactive, data-informed management of buried pipeline infrastructure using explainable AI models with characterized uncertainties.
Quantifying variation in teeth of Late Triassic vertebrates: implications for identification, palaeoecology, and biostratigraphy
Keeble, Emily (Virginia Tech, 2025-12-22)
The Triassic Period (~252–201.5 Ma) represents one of the most dynamic intervals in Earth's history, characterised by major evolutionary radiations following the end-Permian mass extinction, the emergence and diversification of key vertebrate lineages, and the establishment of modern terrestrial ecosystems. Whereas macrofossils tend to capture greater attention, small fossils are equally as important to reconstructing vertebrate communities of ancient ecosystems. This is especially true of fossils found from microvertebrate localities (deposits where 75% of remains are <50mm) as these record both small vertebrates and small parts of larger organisms. Tooth fossils are among the most readily preserved in these sites thanks to their relative hardness and resistance to chemical and physical weathering. However, when teeth are found isolated, it can be difficult to ascribe them to a species or even clade as many teeth come from currently unknown animals, and there is a high degree of convergence with little variation among many archosaur teeth in the Triassic Period. Different methods have been used to quantify variation and identify isolated teeth, particularly within Dinosauria, but 3D methods have been underutilised, despite the increasing availability of scanned fossils and increased information on dimensionality that can be captured by using 3D models. My dissertation quantified variation in tooth morphologies at three different taxonomic levels: within a species, within a clade, and among clades with similar inferred diet, and employs quantitative and qualitative analyses on a broad range of taxa. The goal of this work was to test whether the variation in tooth morphology, once quantified, allows for the identification of isolated teeth of unknown taxa in microvertebrate deposits, and as an aid to determine the biostratigraphic utility of the specimens tested. In my first chapter, I looked at variation in tooth shape and discrete characters within a clade, using aetosaurs (quadrupedal terrestrial pseudosuchians with extensive armour across most of the body) and their close outgroups as an example. Interestingly, isolated aetosaur teeth are exceptionally rare in deposits where other aetosaur fossils, such as osteoderms, are common; however, it is unknown whether this is an artifact of aetosaurs generally having fewer teeth than other archosaurs or the result of the difficulty of identifying isolated aetosaur teeth. To determine if aetosaur teeth are diagnostic and can be identified apart from other Triassic Period archosaurs, I created matrices that use 3D geometric morphometrics (3DGM) and non-metric multidimensional scaling (NMDS) to both examine variation and evaluate whether these techniques could be used to help identify isolated aetosaur teeth, particularly in microvertebrate deposits. I found that, broadly, aetosaurs can be distinguished from the outgroups tested. When isolated teeth suspected to be from aetosaurs were added into the matrix, they plotted within the space of the majority of aetosaurs in the 3DGM. In the NMDS morphospace, they plotted more disparately, but still away from the outgroups, suggesting that these methods are useful in identifying isolated aetosaur teeth. My second chapter utilised the methods from my first chapter and applied them to a wider group of animals – the carnivorous archosaurs of the Triassic Period (e.g. Coelophysis, Batrachotomus, Smilosuchus) to test whether distantly related carnivorous taxa retain a similar tooth form, and whether it is possible to identify teeth to major clade despite these taxa sharing the ancestral tooth shape for this clade (i.e., recurved, serrated, and laterally compressed). I found that even though some taxa (e.g. Diandongosuchus, Ornithosuchus, Riojasuchus) overlap greatly in dental morphospace using 3D morphometrics and within the NMDS analysis morphospace plots, dinosaurs tend to plot away from pseudosuchians in the 3DGM, implying that 3DGM may be useful in separating these two clades. My third and final chapter further addressed intraspecific tooth variation and described a new species of hybodontiform shark based on teeth from a highly productive microvertebrate locality called the 'Green Layer' near the Petrified Forest National Park, Arizona, USA. Chondrichthyan (shark) teeth can be useful for biostratigraphy, but differences in tooth shape throughout the mouth can make disentangling species from one another difficult. Additionally, specimens are also known from a second site within Petrified Forest National Park, and I examined variation between the two assemblages using NMDS and conducted a 3DGM analysis on the 'Green Layer' teeth revealing that there is limited morphological variation in this new species. When I conducted a NMDS analysis, the two assemblages strongly overlapped suggesting that these assemblages are indeed the same species and as tooth shape is highly conserved, they are a useful index taxon for the early Revueltian. My dissertation found that the two methods (3DGM and NMDS) used in each chapter are useful on a broad range of vertebrate taxa in combination, from pseudosuchians to chondrichthyans and highlights the importance of small and isolated fossils in reconstructing vertebrate communities.
Applications of Wastewater-Based Surveillance Within Rural and Unsewered Communities
Price, Sarah Fox (Virginia Tech, 2025-12-22)
Wastewater-based surveillance (WBS) presents the opportunity to describe disease and health trends within human populations both more broadly and at a lower cost than clinical reporting. However, rural communities and 'unsewered' households, the latter of which typically treat household wastewater onsite via septic system, remain a barrier to obtaining data that is truly representative of the population. The degree to which resulting data gaps impact WBS datasets has not been explored. Furthermore, these communities, especially when associated with socioeconomic barriers, can be more vulnerable than the general population to the impacts and circulation of infectious disease and antimicrobial resistance (AMR); this makes them key communities for surveillance and intervention. The goal of this research was to explore existing gaps in rural wastewater coverage in surveillance efforts within the context of AMR. This effort aimed both to systematically characterize prior AMR WBS literature for rural and onsite systems and to examine the occurrence of antibiotic resistance genes (ARGs) within septic systems compared with 'standard' untreated sewage collected at the intake to centralized wastewater treatment plants. Review of the existing literature confirmed that rural, small, and onsite wastewater systems were targets for only a small fraction of WBS campaigns, rendering a meta-analysis of their derived data infeasible. Highly recommended WBS techniques, such as longitudinal sampling campaigns or metagenomic sequencing, were particularly rare for rural applications, suggesting a potential gap in the generalizability of data derived from those techniques. Notably, conventional septic systems, which treat the waste of more than a fifth of the U.S. population, were included in only a handful of prior studies targeting AMR. It should be noted that WBS can be integrated with 'treated effluent' sampling for the purposes of environmental surveillance and risk assessment, which are considered key in the surveillance of AMR given its ability to persist and proliferate in natural settings. Given the reliance of septic systems on environmental systems for final treatment and their association with private drinking water resources, this oversight suggests an area of need for research into potential environmental pathways contributing to AMR spread. Septage (a pumpout of the septic tank contents) and septic tank effluent yielded consistent detection rates of ARGs and comparable ARG concentrations when compared with untreated centralized sewage from similar regions. This suggests, in accordance with the limited prior research, that targeted WBS campaigns including septic systems are feasible. Intriguingly, this also suggests that attenuation of some ARGs in the tank may be limited. However, there were noteworthy differences between the matrices, including a lower abundance of blaCTX-M and higher abundance of intI1 within septage and septic effluent samples compared with sewage; the former represents a gene encoding resistance to clinically-important beta lactam antibiotics and the latter acts as an indicator of multi-antibiotic resistance and propensity to mobilize to new bacterial hosts. However, the general similarity may suggest that centralized sewer collection may sufficiently capture unsewered population health in in some cases.
Analytical and Computational Studies of Lennard-Jones Thin Rods
Wang, Junwen (Virginia Tech, 2025-12-22)
Rod-like particles and structures are ubiquitous in nature and engineered systems. Examples include carbon nanotubes, nanowires, biological filaments, liquid crystal molecules, and colloidal cylinders. Understanding their interactions is crucial for predicting the behavior of these systems at various scales. The Lennard-Jones (LJ) 12-6 potential, one of the most widely used functional forms in molecular simulations, captures both van der Waals attraction and Pauli repulsion between (neutral) atoms. The integrated forms of the LJ potential are frequently used to describe interactions between finite-size objects. While such integrated potentials have been reported for simple geometries such as spheres and planes, compact analytical forms for rod-like particles have remained elusive. This dissertation presents a comprehensive theoretical framework and computational implementation for studying LJ interactions involving thin rod-like particles. Through rigorous mathematical derivations based on Ostrogradsky's integration method, we develop closed form analytical expressions for three fundamental interaction scenarios: (1) rod-rod interaction, (2) sphere-rod interaction, and (3) point particle-rod interaction, all in arbitrary three-dimensional configurations. These analytical forms enable efficient computation of both forces and torques, facilitating their use in molecular dynamics simulations and theoretical analyses. For the rod-rod interaction, we derive an integrated LJ potential for two thin rods of finite or infinite lengths in general skew configurations, including special cases of parallel, crossing, coplanar, and collinear arrangements. Each thin rod is treated as a segment consisting of LJ material points. The expressions encompass the integration of both attractive (1/r^6) and repulsive (1/r^12) components of the LJ potential. We verify these analytical results through comparison with direct numerical integration and explore their physical implications, including the adhesion behavior between rods in various configurations. The sphere-rod interaction potential is derived by integrating the LJ potential between a sphere (treated as a continuous medium) and a thin rod. We obtain compact analytical expressions valid for rods of both finite and infinite lengths. This potential is used to study the adhesion between a sphere and a rod. Interesting scaling relationships are discovered: for sufficiently large spheres and long rods, the equilibrium gap width is approximately constant at 0.787σ, where σ is the LJ length parameter, and the sphere-rod adhesion scales with the square root of the sphere radius. To enable practical applications of these analytical potentials in large-scale molecular dynamics simulations, we implement them as a user package in the Large-scale Atomic/Molecular Massively Parallel Simulator (LAMMPS). In this implementation, the motion of each rod is described through the motion of its two ends, with each end treated as a particle in molecular dynamics simulation. We develop efficient algorithms for neighbor list construction, force decomposition, and coordinate transformation between local and global reference frames. The parallel performance of the implementation is analyzed, demonstrating its capability to simulate systems with thousands of rods efficiently.
Enhanced Soil Moisture and Streamflow Estimation in Ungauged or Data-Scarce Watersheds
Asfaw, Binyam Workeye (Virginia Tech, 2025-12-22)
Watershed models are important tools for quantifying hydrologic processes and evaluating water management practices. These models simplify complex systems through parametrization and assumptions, particularly in semi-distributed frameworks like Soil and Water Assessment Tool (SWAT). Clear conceptualization is critical to realize adequate representation of internal hydrologic processes, particularly in saturated excess runoff dominated watersheds where the spatio-temporal dynamics of saturated areas influence both the distributed (soil moisture state, runoff generation, or pollutant export) and integrated (streamflow) watershed responses. While streamflow is commonly used to constrain model parameters due to its availability and integrative nature, the issue of equifinality, where multiple parameter sets yield similar streamflow, but divergent internal process estimates, can lead to significant uncertainty. This dissertation addresses these challenges by introducing new and improved techniques for representing field scale soil moisture patterns and using satellite soil moisture data for model calibration in ungauged and data-scarce watersheds, especially for watersheds where variable source area runoff mechanism dominate. Chapter 2 proposes the topographic index (TI) as a tool to represent the spatial soil moisture patterns. Using unsupervised machine learning on in-situ soil moisture data from a 4.2 ha field, three distinct soil moisture groups were identified. TI values were classified into three groups using various approaches (equal-interval, equal-area, k-means, Fisher) and digital elevation model (DEM) sources (United States Geological Survey (USGS) and drone-based LiDAR) generating topographic index classes (TIC). Performance was evaluated using Spearman's correlation and misclassification rates. Results showed that low-resolution LiDAR and USGS DEMs outperformed high-resolution LiDAR DEMs, with equal-interval classification yielding the best performance. The result showed resampling to a lower resolution improved the performance of LiDAR DEMs, while TICs derived from publicly available USGS DEMs outperformed those from LiDAR DEMs. Among classification approaches equal-interval provided the highest performance. Overall, the result showed a three classes TIC can represent field scale soil moisture pattern and the effect of DEM type, resolution, and classification approach can be substantial. Chapter 3 develops a variable source area (VSA) SWAT (SWAT-VSA) model for a 14 km² watershed, incorporating the three-class TIC to improve hydrologic response unit (HRU) definition. Downscaled and bias-corrected satellite soil moisture data, along with streamflow data, were used for single and multi-objective calibration. Calibration using soil moisture data improved field-scale moisture estimation, while multi-objective calibration enhanced overall model performance. The three-class TIC structure reduced computational cost while effectively representing VSA dynamics. Chapter 4 demonstrated the utility of downscaled and bias corrected satellite soil moisture for streamflow estimation in ungauged watersheds. Three watersheds in the northeastern US with flow monitoring stations were analyzed. Soil moisture-calibrated models, using root mean squared error (RMSE) and SPatial EFficiency (SPAEF) metrics, were compared to streamflow-calibrated and regionalization-based models. While streamflow data and regionalization technique performed better for streamflow estimation, soil moisture calibrated models incorporating spatial metrics in calibration yielded comparable streamflow predictions. Satellite-based calibration offers an objective alternative to traditional regionalization, especially when properly downscaled, bias-corrected, and scaled. This research demonstrates the potential of satellite soil moisture data to improve hydrologic model performance and reduce uncertainty in data-scarce environments, offering a scalable and objective approach to watershed modeling.
The role of adaptive immunity in Parkinson's pathology following traumatic brain injury
Kelly, Colin Joseph (Virginia Tech, 2025-12-22)
Traumatic brain injury (TBI) increases the risk of Parkinson's disease (PD) development later in life, but the molecular and cellular mechanisms are unknown driving this relationship. A single, mild brain injury can activate both resident and peripheral neuroinflammatory signaling pathways that are similarly activated in the brains of PD patients, likely increasing susceptibility to neurodegeneration. Previous studies have demonstrated that specific subtypes of T cells mediate inflammation in preclinical models of PD in response to the neurotoxic accumulation of alpha synuclein. Certain T cell populations are also known to be activated and recruited to the brain parenchyma at subacute timepoints post-brain injury, and can persist chronically, negatively impacting TBI outcome. Using models of both murine TBI and PD, we evaluated how a pre-existing neuroinflammatory event may exacerbate PD-associated pathologies and behavior. Our transcriptomic analysis of mRNA from purified dopaminergic neurons of mice 90 days post mild TBI (mTBI) revealed upregulation of genes related to neuroinflammation, peripheral immune signaling, and IFN-, in addition to dysregulation of genes known to play a role in PD. Quantification of dopaminergic neurons in the substantia nigra showed significant cell death at 90 days post-injury compared to sham controls, with associated alterations in striatal neurotransmitter levels, like dopamine, leading to behavioral phenotypes. At that same time point, CD8+ T cells are present throughout the brain and around the substantia nigra. When mTBI is induced 30 days prior to induction of PD-associated pathologies via intrastriatal injections of alpha synuclein preformed-fibrils, a similar susceptibility of DA neurons is observed, in addition to an increased severity in alpha synuclein propagation. To examine the role of adaptive immunity in these outcomes, Rag2 KO mice were exposed to the same experimental conditions and displayed significant neuroprotection of the DA neuron population compared to wildtype animals. Taken together, these findings indicate the possibility of a sustained peripheral immune cell infiltration after injury and could support a complex, persistent, and detrimental crosstalk between both resident and peripheral immune cells which negatively affects DA neurons.
Versatile and Sensitive Optical Sensing and Imaging with Spectral Interferometry
Thomas, Joseph Gabriel (Virginia Tech, 2025-12-22)
Monoaminergic Signaling in the Human Brain: New Insights
Hartle, Alec Edward (Virginia Tech, 2025-12-19)
Understanding how neuromodulatory systems cooperate to shape cognitive and behavioral processes remains a central challenge in neuroscience. The monoamines dopamine, serotonin and norepinephrine uniquely contribute to neural computations throughout the forebrain, influencing attention, learning and decision-making. However, resolving these monoaminergic signals at physiologically and behaviorally relevant spatiotemporal scales in the human brain have been constrained by limitations of available techniques. Within this dissertation, I employ a machine learning-enhanced voltammetry (MLEV) technique that can detect sub-second transients of dopamine, serotonin and norepinephrine in the human brain. First, we demonstrate the chemical selectivity of MLEV using optogenetically evoked monoamine release in transgenic mice. We then move away from model organisms and apply MLEV in awake humans while they performed behavioral tasks. In this work, we had subjects play an emotional Stroop task and identified there were differential modulation of monoamines during the presentation of valenced words in the thalamus and anterior cingulate cortex. In a separate experiment, patients with Parkinson's disease or essential tremors played a social reward task. We observed opponency between dopamine and serotonin to positive prediction errors in patients with essential tremor, but not in those with Parkinson's disease. Moreover, patterns of dopaminergic and serotonergic signaling predicted disease state. The work in this dissertation demonstrates that coordinated monoaminergic signaling underlies the computations linking valence and reward processes.
Human-AI Handshaking: Supporting Extreme Sensemaking through Trustworthy Shared Perception
Wilchek, Matthew Ryan (Virginia Tech, 2025-12-19)
Extreme sensemaking occurs when teams of people must build situational awareness in high-stakes, dynamic environments such as search and rescue, military, or security operations. These contexts are marked by uncertainty, fragmented information, and time-critical decisions that stretch human cognitive and physical limits. Artificial intelligence (AI) offers potential assistance, but distributed AI systems that rely on object detection, such as drone swarms, autonomous vehicles, and large-scale sensor networks, face their own challenges, including fluctuating accuracy, false detections, and fragile resilience under real-world dynamics. This dissertation introduces the "Human-AI Handshake", a novel human-AI interaction technique that incorporates human-in-the-loop (HITL) and crowd-in-the-loop (CITL) components to enhance object detection accuracy and shared perception in distributed systems. The Human-AI Handshake addresses core challenges by combining human feedback with AI model performance to mitigate uncertainties and improve trustworthiness. The foundation of the proposed concept comprises four key components: (1) evaluating HITL concepts for improving computer vision accuracy, (2) enhancing shared perception among distributed AI systems to enable better situational awareness, (3) ensuring AI trustworthiness through a synchronized perception trust framework, and (4) interpreting contextual awareness to help AI systems adapt to diverse, real-time scenarios. The technique is tested in extreme sensemaking applications, including augmented reality (AR)-assisted search and rescue, where accurate and reliable object detection is essential. Overall, the Human-AI Handshake provides an assured, scalable solution for improving AI assurance in distributed, dynamic environments, ensuring greater reliability, trust, and performance in critical operations.
Polymeric Melodies: Polyetherimide Modification-Based Acoustic Membranes and Sonochemical Degradation of Polystyrene
Valenzuela, Oscar Enrique (Virginia Tech, 2025-12-19)
The compatibilization of polymers for nanomaterials and the degradation of commodity plastics are two major areas in polymer chemistry. To achieve composite materials that meet the demands of modern technologies, polymers are modified using a vast array of chemistries, enabling the practical and effective deployment of nanomaterials. For example, graphene and carbon nanotubes hold significant promise for the advancement of computational power, energy storage, the reinforcement of commercial and construction materials, and in sensing technologies. However, these materials are often difficult to use independently and must be coupled with a substrate or embedded into a polymer matrix. To this end, several strategies for the generation of polymer/nanomaterial composites have been developed. These approaches have suffered setbacks regarding the full potential of the coupled nanofillers, as the predicted theoretical enhancements often fail to translate from the nanofiller to the matrix. Therefore, it is crucial to research strategies that can unlock the full potential of the materials destined to enable the next stage of technology. In this dissertation, we have modified polyetherimide (PEI) using a crosslinking approach to enable the reinforcement of a thin film membrane via carbon nanotube embedding. In addition to the generation of nanocomposites, effective chemical recycling of waste plastics has become paramount to solving the current pollution crisis. While most see discarded plastic as waste with no inherent value, the modern chemist sees a feedstock waiting to be mined for the valuable moieties and energy within. Judicious treatment of plastic waste can yield fuel sources, pharmaceutical scaffolds, and macromolecular plastics with new properties. However, true material circularity remains elusive due to the need for expensive catalysts, energy-intensive thermochemical processes, and the physical labor involved in collecting waste. In this dissertation, a catalytic, green process is explored to degrade polystyrene (PS) using acoustic energy. PS is degraded using ultrasonication into benzene using substantially lower amounts of catalyst compared to contemporary methods. Thin layers of polymer are often deposited onto single-layer graphene (SLG) to act as support substrates for large radius suspensions of SLG. The thickness of the polymer layer and the functionalities installed onto SLG are controlled to cater to specific applications. However, thin film polymer/SLG composites can be brittle and suffer from rapid degradation upon cycling. The material must be able to withstand a wider range of mechanical use to be commercially viable; therefore, polymer/SLG composites must be mechanically strong and robust. To improve the survivability and overall strength of a PEI/SLG composite, we installed crosslinking moieties onto PEI and introduced carbon nanotubes (CNTs) into the polymer matrix to increase the tensile strength. The CNTs are dispersed using sonication, and the films are spin-coated onto SLG on copper (SLG/Cu) from solutions comprised of functionalized PEI and CNTs in chloroform. The films are annealed to initiate crosslinking, and after removal of the Cu layer, freestanding thin films are recovered. The films show an increase in Young's modulus of up to 5.4 GPa with CNTs, along with a stiffer 2D modulus. The crosslinking of PEI in the presence of CNTs provides a new method for forming covalent linkages between a polymer and CNTs, achieving mechanically robust and reinforced thin films. Plastic waste pollution has only grown as a threat to public health. As plastic production continues to increase, recycling rates have not managed to curtail the inevitable accumulation of waste in the environment. Only 30% of all produced plastic is recycled in developed countries, with nearly non-existent programs in developing nations. While mechanical recycling and incineration enable the physical recycling and energy recovery from plastic waste, the reformed material is often mechanically compromised, and incineration releases toxic fumes into the atmosphere. To address this, chemical degradation and upcycling of plastics have been achieved using various catalysts. However, these processes require high catalyst loadings and intense conditions, lowering the efficiency and economic viability. Ultrasonication can overcome these limitations by leveraging the intense energy release during cavitation to drive a catalytic degradation of PS. Herein, we present the degradation of PS into benzene using ultrasonication with significantly lower molar amounts of aluminum chloride (AlCl3) than is required by thermochemical reactions. The reaction is rapid and able to produce quantitative yields of benzene from PS without any external sources of heat and pressure. Sonication not only requires much lower amounts of energy but also produces benzene in mild conditions, with the reaction proceeding at 0 °C. This work offers a more efficient, truly catalytic approach to the degradation of PS with AlCl3, increasing the economic viability of chemical recycling.
Enhancing Software Maintenance: A Research Investigation on Current Practices, Potential Improvements, and Procedure Automation
Kabir, Md Mahir Asef (Virginia Tech, 2025-12-18)
Software maintenance is one of the most important phases of the Software Development Life Cycle, as it prevents unexpected development issues and ensures long-term reliability. Proper maintenance reduces cost and protects software from security and run-time problems. However, software maintenance is widely recognized as a challenging and resource-demanding phase of the life cycle. Developers frequently encounter difficulties such as managing dependency updates, addressing metadata inconsistencies, and adapting to evolving project requirements. These recurring issues motivate the need for better insights into maintenance practices and the exploration of automated techniques that can improve reliability and efficiency. To address these challenges, this dissertation presents (1) an examination of current maintenance practices by developers, (2) an investigation of the feasibility of using Large Language Models for software maintenance, and (3) the development of a new tool to automatically detect bugs and support automated maintenance. First, we identified security-related best practices for JavaScript developers and examined how well they are followed in open-source projects. Our empirical study of 841 applications revealed frequent violations and showed that developers often ignore best practices due to perceived irrelevance or distrust in tools. These findings highlight limitations in human-driven maintenance and motivate the exploration of automated assistance. Next, we evaluated how Large Language Model (LLM) tools such as ChatGPT perform in maintenance tasks compared to human developers. By analyzing their performance in technical QandA and software revision, we found that ChatGPT provided better answers for 97 out of 130 Stack Overflow questions and successfully revised software for 22 out of 48 maintenance requests. While promising, our results indicate that LLMs struggle with context-specific precision. Finally, we developed a domain-specific language (DSL) and language engine tool (MECHECK) to detect metadata-related bugs in Java programs. We defined 15 rules from Spring and JUnit documentation and evaluated MECHECK using two datasets of 115 enterprise applications. MECHECK detected bugs with 100% precision, 96% recall, and 98% F-score, and identified 152 real-world bugs, 49 of which were confirmed fixes. These results demonstrate that MECHECK helps ensure the correct use of metadata and advances automated software maintenance. In summary, this research provides insight into software maintenance, its challenges, and how the process can be improved from understanding developer behavior, to leveraging AI assistance, to creating automated detection tools.
Modeling the Food-Water Nexus: A Spatio-temporal Accounting of Agricultural Land and Water Use in the United States
Lamsal, Gambhir (Virginia Tech, 2025-12-18)
Agriculture dominates both land and water use in the United States, and plays a pivotal role in both national food security and global agricultural trade. Yet, this critical system faces growing pressures from groundwater depletion, climate change, and competing demands for water across sectors. Addressing these challenges requires an integrated understanding of how croplands and water use have evolved through time, and how more efficient management can support agricultural production without expanding water withdrawals. This dissertation developed a unified, high-resolution framework linking agricultural land use, crop water consumption, and on-farm management to evaluate opportunities for sustainable intensification within the U.S. food–water nexus. This dissertation first created HarvestGRID, a gridded dataset of irrigated and rainfed harvested areas for 30 major crops from 1981 to 2019. Existing datasets often face a trade-off between spatial detail and temporal coverage. Remote sensing provides fine spatial resolution but limited historical depth, while administrative records extend further back in time but have coarse spatial resolution and contain missing values. HarvestGRID bridges this divide by combining USDA survey and census records with satellite-based land-use products to create spatially explicit and temporally consistent maps of harvested area. The dataset captured long-term agricultural shifts, revealing that while the national extent of irrigated cropland has remained relatively stable, irrigation has declined in water-scarce western states and expanded in more humid eastern regions, reflecting adaptive responses to changing water availability. Building on this spatial foundation, this dissertation then created MIrAg-US (Modeled Irrigated Agriculture of the United States), which provided the first multi-decadal, monthly record of crop water consumption for the same 30 crops using process-based crop growth models. U.S. irrigated croplands consumed an average of approximately 154 cubic kilometers of water annually, with about 70 percent derived from blue water sources (irrigation). Alfalfa and corn together accounted for nearly 40 percent of this total, underscoring the dominance of a few key crops in national water demand. Modelled estimates from MIrAg-US were rigorously evaluated against multiple independent data sources, including government water-use records, previously published model estimates, and remotely sensed evapotranspiration datasets. The comparisons demonstrated generally strong agreement, although alignment varied by region and crop type, reflecting both differences in modeling frameworks and input data. Finally, we utilized the modelling framework developed in previous chapters and evaluated the potential for increasing U.S. food production through improved on-farm water management, explicitly accounting for the rebound effect i.e. the reinvestment of saved water into expanded cultivation. Using AquaCrop-OS simulations, we quantified the water savings achievable through the adoption of high-efficiency irrigation technologies (sprinkler and drip) and organic mulching across 13 major irrigated crops, and modeled the reallocation of this saved water within the same watershed. Nationally, these practices could save up to 27.4 billion cubic meters of irrigation water annually (30% of current total applied irrigation), and reallocation of this water could expand irrigated croplands by as much as 6.2 million hectares, primarily by converting rainfed cropland. This expansion would increase national crop production by approximately 21 million metric tons per year (an 8.9% gain), valued at $4.7 billion annually. Together, these studies create a cohesive empirical and modeling foundation for understanding agricultural water sustainability in the United States. Beyond documenting past change, this work establishes a pathway that links crop modeling and human decision-making to guide data-driven strategies for managing water and food systems under a changing climate.

Browse

Recent Submissions