48440 research outputs found
Sort by
Mathematical Modeling, Numerical Framework, and Performance Based Control of Tethered Marine Power Generation Systems
No full textA growing interest in marine renewable energy has driven the development of novel tethered energy-harvesting platforms that operate in fluid environments. This dissertation focuses on the development of high-fidelity dynamic models to enable accurate prediction, stability evaluation, and performance assessment of tethered marine energy harvesting systems operating under complex environmental and operational conditions. The work presented here builds upon baseline operational principles of cross-current motion already incorporated into current technologies such as the Tethered Undersea Kite (TUSK). Controlled cross-current kite trajectories are used to induce kite velocities which exceed the current velocity to enhance generated power. Expanding on these principles, this research presents the modeling, simulation, and control of a recently proposed energy-harvesting platform: the Slaloming Tethered Riverine and Marine Surface (STREAMS) Wing. To rigorously capture the dynamics of STREAMS, principles of computer-aided rigid body dynamics, combined with classical formulation techniques, are used to develop a six-degree-of-freedom (6-DOF) nonlinear dynamic model. This model accurately characterizes the system’s motion and supports reliable performance predictions across a range of operational scenarios. Recognizing that technologies of this kind are typically designed to operate autonomously, The present work focuses on laying the groundwork for the development of robust control schemes that not only enable baseline functionality but also enhance system performance. Specifically, for the STREAMS platform, an event-triggered open-loop control strategy and a Lyapunov-based heading tracking control approach are evaluated. These strategies are intended to establish a foundational framework upon which more sophisticated control methods can be developed and implemented in future research. In addition to the STREAMS dynamics and control, further complexity is introduced through the incorporation of tether dynamics using the TUSK platform as a case study. Here, a framework is developed to investigate how the tether configuration affects overall system performance. The tether is modeled as a compound structure composed of multiple connected elements joined through standard first-order rigid joints. In parallel, the effects of the tether's elastic behavior and mechanical properties are incorporated using a finite element method (FEM)-assisted scheme, which enables the integration of a discretized tether state-space flexible dynamics representation into the existing TUSK with rigid tether dynamic framework. This extended formulation provides a more comprehensive understanding of how tether flexibility and configuration impact the behavior and efficiency of tethered energy-harvesting systems. Design methodologies and control strategies are validated through numerical simulations based on the 6-DOF dynamic models of both STREAMS and TUSK. Sensitivity analyses across various parameter setups not only corroborate the theoretical framework but also yield critical insights into the impact of design and control parameters on system stability, power output, and robustness. Ultimately, this integrated approach bridges the gap between theoretical analysis and practical implementation, paving the way for enhanced system performance and scalability in real-world applications. The proposed models aim to capture the interactions between system components, environmental forces, and operational constraints. For both the STREAMS and TUSK platforms, this includes a detailed treatment of rigid-body hydrodynamics, tether-induced forces, and control-driven responses. Therefore, to provide further insights into the development of these technologies, design and performance considerations are validated through extensive simulations based on a 6-DOF representation of the underwater vehicle dynamics, coupled with tether modeling techniques. In the case of STREAMS Wings these simulations are complemented by parameter sensitivity analyses, in which critical physical and control variables, such as tether attachment location, lift-curve slope, drag coefficients, and control schemes, are systematically varied to identify their effects on velocity augmentation, power generation, stability margins, and control robustness. Moreover, this effort includes performance benchmarking via Froude scaling against commercially available alternatives. For TUSK, tether dynamics are explored under various modeling fidelities to determine their direct effect on the system's performance. Together, these studies not only provide a more realistic motion predictive framework but also provide valuable perspectives for the future implementation of more sophisticated design optimization and control co-design frameworks. This integrated and modular approach bridges the gap between foundational dynamic analysis and real-world system implementation, ultimately paving the way for scalable and efficient deployment of tethered energy-harvesting technologies in complex marine environments
Recycling Plastic Waste and its Education
No full textEvery day, plastic is disposed of to be processed and recycled, however only 9% is actually recycled globally. This waste ends up in landfills, incinerated, or littered, contributing to negative health effects in humans and broader climate change. To educate on proper recycling practices, our team developed a website to encourage students to recycle their plastic waste through a scoring system and a map of campus locations of recycling bins. Through these efforts will demonstrate the ease of recycling and feasibility of an on-campus recycling facility
Stock Market Simulation 2525
No full textThis project was a four-week simulation of the stock market with the intention of comparing two investment strategies: Buy and Hold and Swing Trading strategies. Both strategies were employed with the same portfolio of ten large stocks with $100,000 initial principal of virtual money on the Investopedia simulator. Buy and Hold purchased all shares at the beginning and kept them until the end of the duration. Buy and sell decisions were made based on technical indicators such as RSI and MACD in Swing Trading strategy. Both strategies ended up being profitable, with Buy and Hold returning 12.67% and Swing Trading returning 9.83%. This simulation has better enabled participants to understand stock trading as well as the distinction between passive and active investing methods
UNDERSTANDING HEAT TRANSFER CHARACTERISTICS OF OSCILLATING ELECTROHYDRODYNAMIC CONDUCTION PUMPING IN THE PRESENCE AND ABSENCE OF PHASE CHANGE
No full textThe current trend in electronic systems sees them continue to develop at a remarkable pace and, thus, they become smaller, more complex, and require larger power densities to operate. Similarly, associated thermal management technologies that support them must continue to advance. The existing world of thermal management systems is typically comprised of technologies such as heat pipes, two-phase loops, or phase-changing materials, or steady state thermoelectric coolers to dissipate heat from electronics. These systems, despite their ability to transfer heat, are often met with challenges such as overcoming mechanical deterioration, low efficiency, and are often limited to being passive thermal management methods. Electrohydrodynamic (EHD) conduction pumping is an innovative, low power, and non-mechanical method to pump fluids for thermal management. In addition to these advantages, EHD is also simple in design and is controllable, as it can be integrated into complex and programmable scenarios for optimal thermal management. As a technology, EHD conduction pumping has been proposed as a promising method of thermal management in rigid configurations and uni-directional flows in both space and terrestrial applications. The research conducted in partial fulfillment of the degree of Doctor of Philosophy in Mechanical Engineering expands upon EHD research and directly contributes new experimental and numerical results, new flexible pumping designs and fabrication methods, and brings EHD-driven oscillation to the world of enhanced thermal management. Flexible EHD conduction pumping is particularly relevant to complex geometric conditions, high vibration/shock environments, such as aerospace and space applications, and to wearable technologies. Furthermore, by designing EHD conduction pumping to oscillate flows, enhanced heat transfer is observed when compared to traditional unidirectional pumping due to mixing and convective enhancements, as well as due to influences on bubble departure behavior. The work conducted in this dissertation includes both fundamental and applied investigations. Major results presented are as follows. Firstly, velocity measurements of flexible EHD pumping in various configurations are measured for various working fluids. Next, a novel flexible oscillating EHD conduction pump is tested in a two-phase liquid film to demonstrate the effect of applied oscillation frequency on heat transfer enhancement and critical heat flux. Testing in both saturated conditions and in the presence of non-condensable gases is considered for this study. Particle image velocimetry (PIV) measurements are also taken to capture the flow fields within the liquid film. Next, a numerical evaluation of the effect of oscillation frequency on the heterocharge layer evolution and morphology, flow field, vorticity, and heat transfer characteristics of oscillating EHD conduction pumping in transient and periodic steady state modes in a single-phase liquid channel is presented. After this, a two-phase numerical study is presented, highlighting the effect of applied frequency on the heterocharge layer morphology, fluid flow fields, and liquid/vapor interface in transient and periodic steady state modes. Finally, a two-phase liquid film flow boiling enhancement is demonstrated using EHD and DEP together, raising critical heat flux and sustaining boiling in the presence and absence of gravity. This experiment was flown aboard a parabolic microgravity flight. The cumulative results of this research aim to contribute to the current body of thermal management knowledge by offering active solutions to electronics cooling, and the results continue to support the future of flexible EHD conduction pumping and related thermal management devices for both terrestrial and space applications
Comparative analysis of wavelength-specific UV stress granule formation
No full textStress Granules (SGs) are a type of cytoplasmic biomolecular condensate that form downstream of the integrated stress response (ISR). These membrane-less compartments form by way of liquid-liquid phase separation (LLPS) and sequester various proteins and mRNAs while bulk translation is stalled to conserve cellular energy during stress. Canonically, SGs have been considered dynamic and cytoprotective, allowing cells to evade cell death during stress periods. Canonical SGs are reported to form in response to a variety of different stressors including heavy metals (e.g., arsenic and cadmium), heat shock and viral infection. Upon the end or removal of stress, SGs are cleared via the autophagy pathway and lysosomal degradation. Environmental contaminants, specifically heavy metals, remain a public health concern, yet little is known regarding chronic, physiologically relevant, exposure to heavy metals on the cellular level – specifically in terms of the ISR and SG formation. In Chapter 2 we show that arsenic and chromium induce SG formation in U2OS cells (Figure 2.1). Additionally, we corroborate prior results that chronic low dose arsenic exposure suppresses later acute arsenic-induced SG formation (Figure 2.2). Interestingly, we observed a SG suppression effect when cells were pre-treated with autophagy inhibitor DBeQ (Figures 2.3 and 2.4). Recently, non-canonical SGs have been reported to form in response to a few stressors including nitric oxide, chronic nutrient starvation, and ultraviolet (UV) irradiation. These non-canonical SGs were first branded “non-canonical” based on the absence of a few canonical SG components, including eukaryotic initiation factor 3 (eIF3) and mature poly-adenylated mRNAs. The status quo, as it pertains to SG biology, is the use of fluorescence microscopy to visualize the localization of various proteins at the SG and then infer what stress response pathways may be impacted by the sequestration of those various proteins. Fluorescence microscopy methods are useful however, they are low throughput, population based, time consuming and somewhat biased. There is a need for higher throughput methods to assess SG formation; thus, we set out to develop a novel SG reporter cell line using a split fluorescent protein (FP) plan and mNeonGreen2. In Chapter 3 we report the development of six successfully cloned plasmid constructs (Table 3.1) that facilitate our SG reporter cell line as well as a proof-of-concept transient transfection in Cos7 cells (Figure 3.2). Additionally, we show the generation of three control HeLa T-Rex™ cell lines that will be necessary if and when a SG reporter cell line is developed with a split FP plan and a green fluorescent protein (GFP) (Figure 3.3). The generation of a SG reporter cell line that can be used to quantify SG formation by flow cytometry would be monumental for the field of SG biology and therefore the development of such a cell line should be prioritized. Although we were ultimately unable to produce a SG reporter cell line, we were able to further investigate and compare non-canonical UVB and UVC-induced SG formation. In Chapter 4 we report that UVB and UVC induce robust SG formation in wt U2OS cells (Figure 4.1), however, no UV wavelengths or doses tested in our study induced robust SG formation in the human keratinocyte cell line, HaCaT (Figure 4.5). Additionally, we show that while UVC-induced SGs in U2OS cells appear to be cell cycle dependent, UVB induces SG formation in U2OS cells regardless of cell cycle stage (Figure 4.4). Interestingly, the few HaCaT cells (~10-15 cells out of 250+ cells scored) that did form SGs in response to UVB appear to be exclusively in the G1 phase of the cell cycle, suggesting UVB-induced SG formation is cell cycle dependent in this cell type (Figure 4.5). Our results indicate that there may be stress-specific or cell-type specific differences between UVB and UVC-induced SG formation. Overall, this work adds to the growing pool of literature that aims to understand SG formation and function in response to various chronic and physiologically relevant stressors
Cross-Level Analysis of Side-channel Leakage Assessment
No full textSide-channel leakage assessment (SCLA) identifies unintended information leakage from hardware via power, timing, or electromagnetic emissions. Although traditionally performed post-silicon, pre-silicon SCLA enables early leakage detection through simulation across multiple design abstraction levels, including architecture, micro-architecture, and gate. This thesis introduces a simulation and analysis environment for performing pre-silicon SCLA across these abstraction levels, targeting RISC-V System-on-Chip (SoC) designs. We propose a methodology to identify two types of tracked components: leaky components, whose value changes depending on a Critical Security Parameter (CSP), and value-sensitive components, whose value directly depends on a CSP. Our approach uses both non-specific (e.g., TVLA) and specific (e.g., correlation-based) leakage tests. To support root-cause analysis, we automatically establish connections between components across abstraction levels and clock cycles, enabling the tracing of leakage propagation—even in pipelined designs. We evaluate our approach on the PicoRV32 and IBEX SoCs using benchmark programs and cryptographic operations. The results demonstrate improved leakage analysis capabilities, enabling the identification of both software- and hardware-level sources of leakage, including masking flaws and unintended register interactions in cryptographic implementations
Property Modification for Halide-Based Solid-State Electrolyte through Multiple Element Doping
No full textAll-solid-state lithium-ion batteries (ASSLBs) represent a critical advancement in energy storage technologies due to their potential for high energy density and enhanced safety. Among the various solid-state electrolyte (SSE) options, halide-based SSEs such as Li3InCl6 (LIC) have emerged as promising candidates owing to their high ionic conductivity, chemical stability, and broad cathode compatibility. Nevertheless, challenges persist in simultaneously achieving high bulk conductivity and stable interfacial contact with lithium metal anodes. In this study, we explore dopant engineering as a pathway to address these limitations. Through computational screening and thermodynamic modeling, zirconium (Zr) and iron (Fe) were identified as optimal dopants to enhance bulk ionic conductivity and interfacial stability. The Zr-doped Li2.75In0.75Zr0.25Cl6 achieved a room-temperature ionic conductivity of 5.82 × 10⁻³ S·cm⁻¹, which is the highest reported for halide SSEs. Meanwhile, Fe doping facilitated direct and prolonged contact with the Li anode without requiring a protective sulfide interlayer. The Fe-doped LIC demonstrated stable cycling in symmetric cells for over 200 hours and enabled full-cell operation over 300 cycles with 80% capacity retention. Collectively, this work showcases a synergistic computational-experimental strategy for dopant selection and demonstrates a viable route to engineer halide SSEs with both high conductivity and robust interfacial stability, paving the way for simplified, high-performance ASSLB architectures
Improving LLMs Persuasiveness with Online Debates Data and Argumentation via Multi-Agent AI Systems
No full textDebating ideas can help people learn about, reconcile, or dismiss different perspectives about a theme. LLMs are promising agents for debates as they were trained for conversations using big data, much of it directly collected from the Web. However, existing LLMs can be less persuasive than experienced debaters and generate repetitive statements, which weakens their arguments and can produce overwhelming responses. In addition, LLMs tend to seek the user's approval even if they are wrong, which can make the debate biased towards the user's opinion and arrive at incorrect conclusions. To overcome these challenges, we leveraged fine-tuning from large data and multi-agent AI systems. We fine-tuned a LLama 3.1 Instruct model with 8B parameters using a dataset collected from the subreddit “Change My View” containing arguments from successful debates. For the multi-agent component, we proposed a framework with a summarizer, a planner, and a generator. First, we evaluated the fine-tuned against the pre-trained model in an in-person lab study to determine the relevant debate criteria: conciseness, persuasiveness, and (lack of) sycophancy. After incorporating the human feedback, we developed the multi-agent debater and compared it against the pre-trained model using LLM evaluators. The multi-agent consistently outperformed the pre-trained model in all criteria with even a preference of over 89% in one debate criterion
Improving TCP Slow Start Performance in Wireless Networks with SEARCH
No full textTransmission Control Protocol slow start is designed to gradually ramp up the congestion window, aiming to match the available network capacity while minimizing congestion. However, this mechanism often struggles in high bandwidth-delay product networks, leading to premature exits from slow start or excessive packet loss. This thesis introduces a novel algorithm, Slow start Exit At Right CHokepoint (SEARCH), which aims to address these challenges. SEARCH accurately determines the point at which slow start should exit by analyzing delivered bytes in comparison to expected rates over multiple round-trip times. By smoothing delivery rates over several round-trip times and normalizing bandwidth estimates, SEARCH adapts to varying conditions, preventing early exits while maintaining high throughput and avoiding unnecessary congestion. The thesis details the design and implementation of SEARCH across multiple high bandwidth-delay product network types, including satellite (GEO, LEO), and 4G LTE networks, compared to traditional TCP (CUBIC without HyStart) and default TCP (CUBIC with HyStart). Extensive experiments show that SEARCH consistently exits slow start at the right capacity point, improving throughput and reducing packet loss across diverse network environments
Motion Planning with Guaranteed Problem Space Coverage in Semi-Static Environments
No full textMotion planning in static environments typically involves computing a single feasible path from start to goal, as the environment remains unchanged. In semi-static environments, obstacle positions may vary between queries, but most methods solve each instance from scratch without exploiting this repeatability. Experience-based planners attempt to accelerate planning by retrieving previously computed solutions, but they lack formal guarantees and often fail to generalize to unseen or significantly different obstacle arrangements. Preprocessing-based approaches that offer fixed-time planning guarantees do exist, but they rely on discretization of the obstacle arrangement space and assume uniformity in obstacle geometry—the latter often inducing non-solvable problem instances. In this work, we formalize the notion of problem space coverage in semi-static environments and propose an incremental planning approach that guarantees it. Our method constructs a roadmap by iteratively selecting and solving segregated problem instances to converge toward complete coverage, without the aforementioned limitations. The resulting roadmap ensures that a solution can be retrieved for any valid obstacle configuration without additional planning. We also introduce a verification tool that tests whether a given roadmap satisfies problem space coverage, and returns unsolved problem instances when coverage is incomplete. Finally, we analyze trade-offs in preprocessing time and path quality across varying queries, and demonstrate the benefits of our method over the constant-time motion planning baseline