VTechWorks
VTechWorks provides global access to Virginia Tech scholarship, including journal articles, books, theses, dissertations, conference papers, slide presentations, technical reports, working papers, administrative documents, videos, images, and more by faculty, students, and staff. Faculty can deposit items to VTechWorks from Elements, including journal articles covered by the University open access policy. Email vtechworks@vt.edu for help.
Open Access Policy
Virginia Tech's open access policy enables researchers to deposit the accepted version of scholarly articles with no embargo.
Theses and Dissertations
Virginia Tech was first in the world to require ETDs in 1997, and continues to add scans of older theses and dissertations.
Open Textbooks
More than 50 freely available and openly licensed textbooks are among our most downloaded items.
Communities in VTechWorks
Select a community to browse its collections.
Recent Submissions
Sorption of Perfluorooctanoic Acid (PFOA) in an Oligotrophic Aquifer
Vaidya, Priyashi Prashant (Virginia Tech, 2026-10-02)
Deployment Load Estimation of Composite Tape Spring Spacecraft Booms Using Reverse Engineering and Data-Driven Rational Approximation
Mhadgut, Deven Hemchandra (Virginia Tech, 2026-10-01)
Deployable composite tape spring booms have become ubiquitous for small satellite missions. While these booms offer compact and self-deploying structural solutions, their uncontrolled deployment generates mechanical shocks that threaten delicate spacecraft components. Therefore, understanding the structural dynamics of the booms has been a growing area of interest. As a first step towards deployment-induced shock load estimation, this study characterizes the dynamics of a parabolic composite tape spring boom using an experimental modal analysis and finite element (FE) modeling workflow. This initial work details the testing setup, evaluation procedures, modeling results and uncertainties. To minimize geometric uncertainty, a highly accurate scanned point cloud model was evaluated against standard discrete cross-section models. Furthermore, the FE model based on the point cloud approach closely matched the experimental natural frequencies and mode shapes of the first four fundamental resonant modes, demonstrating a significant improvement in accuracy compared to the idealized constant cross-section baseline. However, the results indicate that further refinement of boundary condition modeling could improve correlation for higher-order modes, providing a valuable direction for model improvement.
Subsequently, advanced data-driven modeling techniques, including vector fitting and the parametric Adaptive Antoulas-Anderson (p-AAA) algorithm, are used to analyze experimental frequency response functions. By constructing numerical models directly from experimental frequency responses, these methods bypass explicit modeling of physical parameters necessary for high-fidelity FE models. These data-driven models are validated using test signals and then used to estimate the deployment loads of the boom using transfer function inversion. Experimental testing revealed load-dependent nonlinearity where the structure's fundamental bending mode frequency shifted with an increase in input load amplitude. Cross-validation demonstrated poor predictive performance when a model trained at a specific load level was applied to different loading conditions, proving that non-parametric methods lack the generalization required for varying dynamic loads.
Addressing these limitations, a parametric modeling framework using the p-AAA algorithm was implemented and compared to the non-parametric load-specific models. Experimental validation was transitioned to a shaker table configuration with closed-loop control to better replicate end-of-deployment shocks by applying model-predicted forces. During physical deployment-induced impulse response replication, the vibration control system successfully matched the target acceleration profiles in both the time and frequency domains. The output velocity response also closely matched the actual deployment velocity response of the boom. However, isolating the parametric model from hardware limitations revealed that fundamental differences between the constant amplitude sine-sweep testing and transient deployment dynamics still present challenges in perfectly replicating the physical events.
Finally, the effects of temperature on deployment dynamics, including changes in velocity, deployment time, and natural frequencies, are investigated through experiments conducted in thermal vacuum and ambient conditions. While the physical testing was constrained to a sparse dataset of discrete temperature points, the parametric model successfully captured the system's behavior within this available data. However, internal holdout testing results highlighted the need for additional temperature data points to improve interpolation performance of the model. By discussing these model limitations and exploring future validation steps, this work emphasizes the critical role of data-driven techniques in understanding the effects of uncontrolled deployments in the space environment.
This work advances the predictive modeling of deployable space structures through three primary contributions. First, it quantified that mitigating geometric uncertainty via 3D scanning is essential for accurate modal analysis, reducing frequency errors to under 5% for critical modes. Second, it successfully established a transfer function inversion framework while demonstrating the need for parametric modeling of load-dependent, nonlinear dynamic behaviors across continuous inputs. Finally, it performed closed-loop experimental validation, confirming the parametric model's potential to accurately reconstruct input forces directly from dynamic response data for on-orbit deployment diagnostics.
Advancing Blockchain-Oriented Software Engineering: Empirical Investigations and AI-Augmented Tooling for Secure Smart Contract Development
Khalid, Shawal (Virginia Tech, 2026-10-01)
Blockchain ecosystems are evolving rapidly, placing increasing demands on developers who build decentralized applications, including smart contracts—self-executing programs deployed on the blockchain that enforce agreements without intermediaries. Blockchain developers face immutable code, high-stakes systems, rapid cycles, and social media pressures—challenges often unmet by traditional software engineering.
This dissertation investigates how blockchain developers work, how their behaviors are shaped by external signals—such as textit{crypto signals}, defined as public social media messages (e.g., tweets) from influential figures that affect market sentiment—and how generative AI tools can support secure smart contract development. Through a series of research studies, we: (1) surveyed blockchain developers to identify software engineering practices and pain points; (2) analyzed the influence of crypto signals on user activity and GitHub development patterns, and (3) evaluated the capabilities of large language models (LLMs), including ChatGPT, Gemini, and ChainGPT, for Solidity smart contract generation; and (4) evaluated smart contract security analyzers across two datasets, revealing substantial differences in vulnerability coverage, false-positive burden, execution robustness, and cross-dataset behavior.
Building on these findings, this dissertation designs, implements, and evaluates textit{ChainSentinel}, an evidence-oriented framework for smart contract auditing. ChainSentinel integrates heterogeneous security analyzers within a provenance-preserving workflow that reconciles and prioritizes findings and supports selected executable validation and repair-oriented re-evaluation. Across the evaluated datasets, the framework substantially reduced the burden of unfiltered analyzer aggregation while continuing to produce audit reports under partial analyzer failure.
Collectively, this work advances blockchain-oriented software engineering by connecting empirical understanding of blockchain development, generative AI, and security-analysis tools with the design of evidence-oriented techniques for more secure and transparent smart contract development and auditing.
Deep Lunar Rock Assessment with Vision and Vibration
Ruan, Yiyan (Virginia Tech, 2026-10-01)
Autonomous lunar site preparation requires assessing individual rocks before excavation or relocation. Strong illumination contrast, occlusion, and burial make this difficult: observations from different viewpoints must be assigned to the same rock before its geometry can be estimated, while surface observations alone provide limited information about embedment. This dissertation develops visual methods for recovering rock identity and geometry, complemented by a separately evaluated vibration-based method for estimating burial depth.
First, an RGB-plus-inverted-depth segmentation model identifies individual rocks, and calibrated projection converts their masks into metric three-dimensional surface observations. Second, ShadowCorr determines which observations belong to the same rock by projecting volumetric shadows behind visible surfaces and learning their shared spatial support. This approach does not require appearance similarity or direct surface overlap. Separate synthetic evaluations achieved average correspondence purity of 99.1% with complete masks and 98.9% with controlled mask erosion.
Third, the associated observations support geometric completion using a covariance-ellipsoid prior, followed by mesh construction and volume and centroid estimation. In the four-view synthetic evaluation, matched, valid reconstructions achieved a median volume error of 16.5% and a median centroid error equal to 4.0% of the true longest rock dimension. Compared with visible-point convex-hull and alpha-shape closures, completion improved centroid accuracy while producing higher volume error.
Fourth, impulse excitation and scanning laser Doppler vibrometry measure the response of partially buried specimens. A neural-network model uses vibration features and exposed geometry to estimate burial depth without measured rock mass or soil parameters as inputs. Separate laboratory experiments with 18 concrete specimens in sand achieved a root mean squared depth error of 0.94 cm under nested leave-one-rock-out evaluation, with each specimen excluded from model development when tested.
Together, these studies establish methods for recovering individual-rock geometry from incomplete visual observations and obtaining complementary burial-depth information from vibration. The results support future robotic assessment before rock removal; validation with natural rocks and lunar conditions, automated measurement, and integration of the sensing methods remain future work.
Thirst
Kapoor, Karan (Virginia Tech, 2026-10-01)
Thirst is a collection of poems examining the speaker's relationship with his alcoholic father, and the paradox of writing toward a subject who resists being written about. Moving through ghazals, villanelles, prose poems, questionnaires, and fragment poems, the collection treats formal restlessness as an ethical stance — the speaker's incomplete knowledge of his father is at the heart of the book's narrative structure. Recurring motifs — whales, wolves, mirrors, birds, and the gut as a central bodily site — bind the poems into a unified imaginative world. Drawing on influences including Leonard Cohen, Bob Hicok, Alejandra Pizarnik, Franz Wright, Agha Shahid Ali, and Mohammed Rafi, Thirst argues that form must enact, rather than describe, the conditions of its making: to write toward a resistant father is to inherit his silences and to build, from refusal, a language for what cannot be said directly.