137091 research outputs found
Sort by
Effects Of Reactant Inhomogeneity On Lean-Premixed-Prevaporized Gas Turbine Combustor Performance
No full textWith the growing interest in lean-burn combustors for aviation gas turbine engines, further understanding of the connections between combustor stability and pollutant emissions is required to achieve reliable and sustainable combustor designs. An important parameter that remains to be systematically studied in regard to lean-premixed-prevaporized (LPP) combustion is the level of fuel-air prevaporization and premixedness. Deviations from perfect premixing, henceforth referred to as reactant inhomogeneities, have been shown to enhance both laminar and turbulent flame stability through partial premixing and stratification phenomena. Unfortunately, this comes at the cost of increased pollutant emissions, such as nitrogen oxides (NOx), due to poorly mixed “hot spots” which produce disproportionate amounts of pollutants compared to the combustion of a homogeneous mixture. Studies of reactant inhomogeneity phenomena have been conducted on benchtop burners with canonical geometries operating with gaseous fuels at atmospheric pressures. This dissertation explores when and how reactant inhomogeneity effects influence a liquid-fueled LPP combustor operating at flight relevant conditions. A multi-element, high pressure LPP test facility was constructed featuring a variable premixer, enabling fuel injection at one of four locations, each a different axial distance from the combustor inlet. This premixer enabled the systematic study of the effects of reactant inhomogeneity on LPP combustor performance. NOx emissions were shown to increase with increased reactant inhomogeneity, demonstrating the significant effect of hot spots on pollutant production. However, the effect of reactant inhomogeneity on lean blowoff limits showed conflicting trends, with results suggesting that optimal stability in the LPP combustor was not achieved with perfect premixing. Through the development and deployment of advanced laser diagnostic techniques, the mechanisms through which reactant inhomogeneity influenced combustor performance were studied, the results of which further supported the hypothesis that an optimal level or reactant inhomogeneity exits for LPP combustor performance
Stochastic Motion Planning and Control under Uncertainties
No full textRobotic systems operating in real-world environments must make decisions under pervasive uncertainties arising from imperfect models, sensor noise, actuator errors, and external disturbances. This dissertation develops a unified probabilistic framework for decision-making, motion planning, and control under uncertainty, grounded in stochastic optimal control and probabilistic inference.
At the core of the formulation is a stochastic optimal control problem for general nonlinear and hybrid dynamical systems with nonconvex cost functions. This formulation encompasses a wide range of robotic tasks while revealing two central challenges: (i) the computational intractability of solving stochastic control problems over high-dimensional trajectory distributions, and (ii) the need for algorithms that achieve both performance optimality and robustness to uncertainty.
To address these challenges, this dissertation introduces several novel methods:
(1) Gaussian Variational Inference Motion Planning (GVIMP), which frames motion planning as a variational inference problem in the space of trajectory distributions, providing a principled approach to approximate stochastic optimal control;
(2) Parallel Gaussian Variational Inference Motion Planning (P-GVIMP), an efficient proximal extension of GVIMP that exploits sparse factor-graph structures and Gaussian Belief Propagation (GBP) for GPU-parallelized gradient computation, enabling scalable planning for nonlinear stochastic systems;
(3) an iterative covariance steering framework based on proximal gradient methods and iterative linearization, providing closed-loop solutions for nonlinear stochastic systems with prescribed uncertainty boundary conditions; and
(4) two complementary algorithms for hybrid stochastic systems—Hybrid Covariance Steering (H-CS) for linear stochastic flows and Hybrid Path Integral Control (H-PI) for nonlinear flows—both derived from a unified path-distribution control formulation.
Together, these contributions establish a cohesive theoretical and computational foundation that bridges stochastic optimal control, variational inference, and hybrid dynamical systems. The proposed methods are validated across diverse robotic benchmarks, demonstrating robust and efficient motion planning under uncertainty. Future directions include receding-horizon extensions for hybrid systems and learning-based multi-modal variational motion planning
Restorative Places in Urban Spaces: Searching for Synthetic Affordances to Foster Equitable Access to Affective-Cognitive Restoration
No full textEnvironmental psychology literature supports the importance of nature for meeting the affective-cognitive restorative needs of people, that may then enable them to address higher-level personal needs such as those identified by Maslow (1943/1954). However, understanding is less clear about what synthetic (“man-made”) features of urban environments might also offer some degree of personal restoration if configured properly. Improving this understanding may be important to support inner-city neighborhoods, where residents may not be able to easily access nature, or address their higher-level needs equitably with their suburban counterparts. The present study seeks to address this gap in understanding by employing an immersive virtual reality (IVR) process, using a psychological survey and physiological measurements (electrocardiogram (EKG), galvanic skin response (GSR), and photoplethysmography (PPG)), to assess affective-cognitive restoration in 73 volunteer participants. Each participant was exposed to thirteen 60-second, 360-degree stereoscopic IVR videos that included spatial audio. Statistically controlling for nature and other environmental factors, the results for the physiological measurements were inconclusive, but three psychological dimensions (emotional, cognitive, behavioral) provided multiple key findings. One key finding revealed that a balance in complexity is required between visual-traffic urban complexity and copresence (a “people complexity”), as well as with other synthetic affordances, such as a space’s visual shape. This balance appears to address human limitations in cognitive attentional resources as discussed in key environmental psychological literature (e.g. Kaplan & Kaplan, 1989). This has implications for how urban planners, designers, and architects may design urban spaces to be more restorative. Another key finding showed that each of the psychological dimensions have slightly different restorative needs from their urban spaces. These needs include less complexity and fewer people for the emotional dimension, but additional people and more complexity (up to a point) to support the cognitive dimension. The findings also indicate that higher income and higher social capital have a relationship with more restorativeness, such that urban neighborhoods with lower income levels and lower social capital may not have equitable access to restorative urban places compared to higher-level areas. Overall, of the tested synthetic features, urban spaces that are in better-condition and better-maintained, with more visual convexity, and with a balance of complexities from visual, traffic, people presence and other synthetic features, depending on the pursued psychological dimension, appear to be the most conducive to support people’s daily restorative needs in a predominantly urban environment. Together, these findings provide city leaders, designers, and citizens with evidence to refocus resources toward a more-equitable and sustainable urban development—using the lens of affective-cognitive restoration
Novel techniques for targeted DNA diversification and directed evolution in Saccharomyces cerevisiae
No full textDirected evolution is a powerful strategy for enhancing genes and proteins. This design paradigm involves creating diversity in genetic sequences in vitro and then selecting for improved function in a host. A desirable extension of these techniques is in vivo or continuous diversification, which would eliminate many laborious cloning steps typically involved in directed evolution. Developing B cells naturally carry out a process of continuous directed evolution of antibodies by applying two main processes: V(D)J recombination and somatic hypermutation. However, B cells are challenging to culture and manipulate. Therefore, we engineered a more tractable, eukaryotic organism, Saccharomyces cerevisiae, to carry out V(D)J recombination or somatic hypermutation and apply these customized yeast for in vivo mutagenesis and evolution.
First, to implement V(D)J recombination in yeast, we integrated the key murine recombination activating genes, RAG1 and RAG2. Using fluorescence microscopy, we identified that RAG1 has poor nuclear localization but that this could be overcome by truncating the protein. By using a novel split antibiotic resistance assay with homology-directed repair, we demonstrated that yeast can make coding joints after RAG cutting. We increased the rate of our yeast’s assisted recombination by over 500-fold by employing codon optimization, co-expressing the HMGB1 DNA-binding protein, improving protein nuclear localization and expression, and evaluating RAG1 truncations. We further showed that our platform could assay the severity of several disease-causing RAG1 mutations. Finally, we used our engineered yeast to simultaneously generate up to three unique fluorescent proteins or two distinct antibody fragments starting from an array of nonfunctional gene segments.
Second, to recapitulate somatic hypermutation, we developed and optimized a CRISPR diversifying base editor for yeast (yDBE). This system functions in vivo and utilizes a dCas9-directed cytidine deaminase to diversify DNA in a targeted, rapid, and high-breadth manner. To develop yDBE, we enhanced the mutation rate of an initial base editor by employing improved deaminase variants and characterizing several aptamer-embedded guide RNAs. By performing high-throughput sequencing, we showed that our base editor enables a mutation rate of up to 8.4×10-4 substitutions/bp/generation over a window of 100 bp. We then demonstrated the ability of the yDBE platform to improve the affinity of a displayed antibody fragment. Next, we built gRNA-tRNA arrays that could express up to six gRNAs simultaneously and used them to target up to three loci at once. Lastly, we applied the base editor to engineer noncoding DNA, creating promoters with both increased and decreased transcriptional activity.
By following B-cell antibody diversification, we have developed two complementary tools for in vivo directed evolution in yeast. Each can facilitate a variety of directed evolution experiments for both antibodies and almost any user-specified gene.Ph.D.Chemical and Biomolecular Engineerin
Multiscale Characterization and Design for 3D Concrete Printing: Material Formulation, Printing Process, and Mechanical Behavior
No full textThe field of 3D concrete printing (3DCP) has attracted growing interest from both research and industry, yet its successful implementation remains limited by an incomplete understanding of the processes governing printability and structural performance. While prior work has emphasized the importance of fresh-state material tailoring and process design and control, the fundamental mechanisms linking these criteria to material behavior and mechanical performance are not yet well established. This report investigates the relationships between 3DCP processing, material structure, and mechanical response in cementitious materials. The influence of printing is explicitly examined by linking microstructural characteristics of printed and cast components to their macroscopic physical behavior. Rheological measurements are used to guide the development of printable material formulations and to evaluate the role of fresh-state mechanical behavior in printing success. Additionally, the compressive and flexural responses of printed and cast specimens are quantitatively and qualitatively compared to assess the effects of printing-induced anisotropy and interlayer interfaces on mechanical failure. Collectively, this work advances the scientific understanding of cementitious materials in the context of additive manufacturing, contributing to improved reliability and performance of 3D concrete printing as a construction technology
Design and Development of an Anatomically Inspired Compliant Palm to Adaptively Reconfigure Precision and Power Grasps
No full textThe interchangeability between using various grasping configurations such as precision and power is proposed to be a key factor in augmenting a robotic hand’s functionality in various applications. One type of grasping is distinct from another due to power grasping’s added dependence on the palm. The design and integration of an equivalent palm in a robotic hand was created through an anatomically inspired process. Conducting a study tracking the movements of markers on a human palm, simplifying the complex behavior of bone, skin, and muscles was reduced to modeling a palm in the robotic hand as a series of interconnected hyper-elastic compliant beams (CB). Simulation of this Compliant Anatomical Palmar Mechanism (CAPM) found similar shaping characteristics in humans during grasping. Contact forces are mapped across the surface of a spherical object to estimate the contact regions between precision and power type grasping. By comparing these two grasps achieved by the same hand, a modeled task of static grasping of a spherical object subjected to an external force measures the effectiveness of the integration of a palm to stabilize grasps and withstand larger amounts of torque in specific applications. These findings demonstrate the potential of incorporating the palm as an essential mechanism during grasping to transition from precision to power to withstand unpredictability in forces in its operating environment. Adapting between these two configurations, the term “efficient grasp switching” describes this potential need for hardware adjustment that redefines the way robotic systems can be evaluated to perform both precision and power grasping.Ph.D.Mechanical Engineerin
Engineering Multivalent Nanobodies Against Amyloid Proteins and Other Antigens
No full textTauopathies, such as Alzheimer’s disease, are neurodegenerative diseases that involve the misfolding and deposition of aggregates of the amyloid protein tau. These diseases are among the most widespread neurodegenerative diseases, yet there are few safe and effective disease-modifying treatments for them. Conformational antibodies and antibody fragments that target the various aggregate forms of tau are promising candidates for treatments to slow the progression of these tauopathies and are useful as reagents to understand the aggregation of tau and its role in disease progression. In this dissertation, we have explored the in vitro development and characterization of multivalent nanobodies, or single-domain antibody fragments, targeting complex and heterogeneous tau aggregates.
We began with the development of a simple approach using a synthetic yeast surface display nanobody library and in vitro cell sorting to identify a pan-tau nanobody with specificity for tau protein relative to other amyloid proteins. We have shown that multivalent versions of our lead tau-binding nanobody have increased avidity towards tau aggregates and recognize pathogenic tau found in the brains of tau transgenic mice. We have also characterized tau fibril-specific nanobodies and modified them for improved delivery past the blood-brain barrier. Next, we modified our sorting strategy to generate conformational nanobodies that target oligomeric tau, a form of aggregated tau which is suspected to be the most toxic form present in Alzheimer’s disease. We demonstrated that our nanobodies are specific for tau oligomers relative to tau monomer and fibrils and bind to tau oligomers in brain samples from Alzheimer’s disease patients. We have extended this work to screen for nanobodies that target oligomers of another amyloid protein involved in Alzheimer’s disease, amyloid-β. Finally, we applied our in vitro antibody discovery strategies to target an antigen involved in infectious disease—the spike protein of the SARS-CoV-2 virus. We created multivalent nanobodies that bind with high affinity to the XBB spike protein and provide protection against an XBB challenge in mice.
Overall, this work demonstrates significant progress in the development and characterization of nanobodies specific for complex multimeric antigens. The multivalent nanobodies that we have generated can be used to study amyloid proteins and their involvement in neurodegenerative disease progression and can be further engineered into potent therapies for neurodegenerative or infectious diseases.Ph.D.Bioengineerin
Comparative Analysis of Traffic Sign Tracking Methods: From Classical Trackers to Temporal YOLO Extensions and Trajectory Optimization
No full textEfficient monitoring of roadway infrastructure is essential for safety, maintenance planning, and regulatory compliance. Traditional traffic sign inventory methods rely on manual field surveys or static imagery, which are costly and difficult to scale. Mobile video acquisition enables continuous capture of roadway environments but introduces the challenge of maintaining consistent identities for traffic signs across motion blur, occlusions, and large viewpoint changes.
This thesis investigates traffic sign tracking in dash-cam video as part of a larger detection–tracking–classification–localization pipeline deployed for the Pima County roadway asset inventory project. The study benchmarks classical motion-based trackers, modern association-based algorithms, and a proposed YOLOv11-based tracker augmented with an appearance-embedding head. Two post-tracking refinement modules—CoTracker for reconnecting fragmented trajectories and a position-based graph smoothing method for identity cleanup—are also evaluated.
Experiments on a curated, frame-accurate annotated dataset show that the proposed method improves identity stability and reduces fragmentation relative to widely used baselines. The results demonstrate that lightweight appearance-aware tracking, combined with targeted post-processing, provides an effective balance between accuracy and computational efficiency for large-scale roadway asset inventory applications
Metric x Design: Overcoming Metric Integration Barriers in Design Process Through Toolkit Development
No full textIntegrating metrics into design practice has become increasingly important as organizations seek to balance data-informed decision-making with creative innovation, driven by the vast amounts of data generated by daily users and consumers. Despite this trend, designers face significant challenges incorporating quantitative measurement into traditionally qualitative processes, with few practical frameworks available to support this integration. This study developed and evaluated the Metric x Design Toolkit, designed to empower designers to meaningfully participate in data-driven decision-making while maintaining the value of qualitative design methods. Subject Matter Expert interviews were conducted, and the toolkit was developed following the Double Diamond framework for design and innovation. The evaluation included a workshop with pre/post assessments to identify four primary integration barriers: conceptual understanding, practical application, contextual learning environments, and attitudinal resistance. Workshop results showed measurable improvements in designers' metric competencies, including increases in knowledge (+1.33), confidence (+1.78), and preparedness (+1.45) on a 5-point scale, based on pre/post-survey averages. The Metric x Design Toolkit achieved a System Usability Scale score of 83.6, indicating excellent usability based on the SUS adjective rating scale. These findings demonstrate that the systematically designed Metric x Design Toolkit effectively bridges the qualitative-quantitative divide in design practice by complementing established design methods with quantitative insights through metric integration, offering a theoretical framework for understanding metric integration challenges and a practical solution for applying measurement in the design process.M.S.Industrial Desig
Optimization and Capacitive Interface Design for a Resonant MEMS Embedded Chemical Sensing System
No full textThere is a clear need in multiple industries for a cost-effective and portable sensor that is both selective and sensitive. The primary objective of this thesis is to optimize the design of a battery-powered embedded system that utilizes a MEMS-based (microelectromechanical systems) resonant sensor to detect volatile organic compounds (VOCs), enabling real-time data collection and processing in the field while being small enough to be wearable. This new, portable, system is compared to the benchtop closed loop system previously used in the iSenSys lab, and its advantages are discussed.
Extensive testing and characterization were first conducted on a previously designed portable automatic gain control (AGC) system in the lab. After this characterization was complete, a new iteration of the system with additional software and features was designed and tested with multiple VOCs. The data collected from these tests are compared to the benchtop closed loop system to confirm that the new design operates as expected while remaining portable. All of the features were verified and added to a new version of the board.
The iSenSys MEMS sensor also has electrodes on its surface to perform capacitance sensing, but no portable interface for this yet exists. Measuring capacitance would provide a dataset independent of frequency to help identify analytes, improving the selectivity of the device. Part of this project aims to accomplish this by developing an effective and portable interface using a commercial capacitance to digital converter. A prototype was constructed using an Arduino to test the viability of this interface and the performance and sensitivity of capacitance sensing on an embedded system is quantified and analyzed.
Moving forward, a new board that combines the closed loop frequency sensing with the new features and capacitive sensing interface was designed, fabricated, and assembled. This system includes multiplexors, allowing it to collect data from four devices on each sensor die instead of only being connected to one device like the previous boards. Preliminary testing for basic operation has been completed on this board, but further testing with multiple gases is still in progress.M.S.Electrical and Computer Engineerin