8413 research outputs found
Sort by
Advancing Multi-Axis Force/Torque Sensing via Beam Optimization, Self-Decoupling Mechanisms, and Morphing-Based Mechanical Intelligence
No full textThis dissertation advances multi-axis force and torque sensing by unifying beam-optimized compliant mechanisms, 3D structural self-decoupling, in-measurement morphing–based mechanical intelligence, and a novel six-axis self-decoupling architecture into a cohesive framework, establishing a new paradigm for high-performance force/torque measurement. The overarching research vision is to develop a class of adaptive, task-aware F/T sensors capable of dynamically reconfiguring their mechanical properties to maximize measurement accuracy, robustness, and operational safety. This work establishes the scientific and engineering foundations for mechanically intelligent sensing systems—devices that not only measure applied forces but also actively adapt to interaction conditions, protect themselves from overload, and optimize information delivery across diverse environments. The dissertation first presents a cross-structured 3D force sensor optimized via parametric beam design and finite-element analysis. Examining various beam geometries and slot dimensions reveals how local strain distributions affect sensitivity and parasitic coupling, leading to an arc-shaped, double-layer rectangular beam configuration that enhances bending sensitivity, reduces cross-talk, and remains manufacturable. Building on these insights, a soft 3D force sensor is developed using a hollow square-column architecture with embedded piezoresistive films and a modified Wheatstone-bridge interface. This compact, volume-efficient design eliminates the need for a rigid frame, confines deformation to the free-end beams, and maximizes strain utilization. Symmetric placement of the sensing films combined with the modified bridge configuration provides inherent mechanical and electrical self-decoupling, reducing inter-axis interference and partially compensating for temperature-induced drift. To address nonlinear and history-dependent behaviors, a generalized Preisach hysteresis model and its inverse are formulated, relaxing classical assumptions and identifying a two-dimensional density function directly from experimental data. This approach accurately reproduces measured force–voltage loops, and the inverse model substantially mitigates hysteresis, improving both accuracy and repeatability of the reconstructed force signals. The dissertation further introduces a mechanically intelligent 3D force sensor with morphing cantilever beams, enabling real-time reconfiguration of structural compliance. By switching among discrete morphing states—each corresponding to a different effective beam length and deformation mode—the sensor achieves variable stiffness, tunable sensitivity, and adaptive multi-axis flexure. Compliant states support high-resolution measurement of small forces, whereas stiffer states provide large-load tolerance and intrinsic overload protection, overcoming the traditional trade-off between sensitivity and robustness. Morphing mechanics are co-designed with self-decoupling bridge circuits to maintain low cross-axis interference across all configurations, preserving a well-conditioned mapping from forces to sensor outputs. This integration produces a compact, mechanically intelligent sensing system combining adaptive stiffness, morphing mechanics, tunable sensitivity, and self-protection. Finally, as a step toward six-axis force/torque sensing, a new self-decoupling mechanism is proposed. This mechanism extends the structural and circuit principles to all six components of force and torque, demonstrating the feasibility of mechanically intelligent, self-decoupling architectures for six-dimensional measurement and providing a clear pathway for future refinement, implementation, and enhanced functionality. In summary, this dissertation establishes a comprehensive framework for next-generation multi-axis F/T sensors that are mechanically intelligent, self-decoupling, and adaptively tunable. Across rigid, soft, and morphing architectures, it demonstrates how structural optimization, compliant mechanisms, and morphing-based adaptivity can be co-designed to deliver high resolution, broad dynamic range, and robust operation. These contributions lay the foundation for a new class of task-aware force sensors with potential applications in dexterous robotics, surgical and haptic devices, wearable and prosthetic technologies, precision manufacturing, and human–robot collaboration—domains where high-fidelity, adaptive, and overload-tolerant force sensing is critical
Nuclear spin coherence times of HD trapped in solid parahydrogen matrices
No full textSolid parahydrogen is an appealing host material for nuclear-spin spectroscopy and precision measurements, providing a weakly interacting environment for embedded molecules, and a high degree of magnetic purity. While previous studies have demonstrated the potential of doped solid-parahydrogen matrices for high-resolution infrared spectroscopy, systematic measurements of nuclear spin coherence and relaxation across controlled impurity conditions remain limited. This dissertation describes the design, construction, and operation of an apparatus that addresses the need to improve existing coherence-time measurements. In this dissertation, we outline the scientific motivation for measuring nuclear spin coherence times. We then detail the apparatus we constructed, which integrates a two-stage orthohydrogen-to-parahydrogen converter, a cryogenic vapor-deposition system, and a high-field NMR spectrometer designed for coherent spin excitation and detection. Using this platform, we carry out free-induction decay spectroscopy, multi-Lorentzian lineshape analysis, spin-echo measurements, and longitudinal relaxation studies on parahydrogen samples doped with HD molecules. Across a range of sample conditions, we observe an improvement on transverse relaxation times over previous studies, and correspondingly measure long longitudinal relaxation. We probe samples with orthohydrogen fractions three orders of magnitude lower than in prior work. In low-impurity samples, the resulting spectral resolution is sufficient to resolve the -coupling hyperfine structure in HD, demonstrating the capability of the apparatus to access narrow-linewidth features in parahydrogen solids. We conclude with a summary of coherence time results and a discussion of future efforts to improve our sensitivity. The platform developed in this work provides a foundation for future spectroscopy and precision-measurement experiments involving molecules in solid parahydrogen
STREAmSML: A GPU-enabled platform for sensing, modeling and control of compressible turbulent flows
No full textHigh-speed aerodynamic flows such as supersonic turbulent boundary layers and shock boundarylayer interactions (SBLI) pose significant challenges for flow control due to strong compressibility
effects, shock-induced separation, and nonlinear multiscale turbulence. At the same time, rapid
advances in machine learning have created new opportunities for data-driven sensing, modeling,
and control; provided that standardized, high-fidelity simulation platforms exist to support repro-
ducible evaluation across algorithms. This thesis introduces STREAmSML, a GPU-accelerated,
Gymnasium-compliant platform built upon the validated STREAmS direct numerical simulation
(DNS) solver, enabling systematic investigation of sensing, reduced-order modeling, and closed-loop
actuation for compressible turbulent flows at supersonic Mach numbers.
STREAmSML integrates a modular Python–Fortran/CUDA architecture that permits reinforcement-learning-based, classical, and open-loop jet actuation strategies while preserving the underlying solver’s numerical accuracy and physical fidelity. A span-averaged observation framework reduces the dimensionality of DNS data by two orders of magnitude, retaining the large-scale coherent
structures essential for control while enabling tractable neural-network-based policies.
Using this platform, sparse sensor placement strategies are first developed for both boundary-
layer and SBLI configurations. Interpolatory and greedy algorithms (QR, TPGR) achieve low recon-
struction error with relatively few sensing locations. Reduced-order models based on Dynamic Mode
Decomposition with control (DMDc) are then constructed. For the boundary layer, moderate POD
truncation (90–95% energy) yields superior predictive performance by removing noise-dominated
high-order modes. For SBLI, the flow exhibits a strongly low-rank structure: the first 4–6 energetic
modes capture approximately 90% of the total variance, enabling compact reduced-order models
that isolate dominant separation-bubble and shock-motion dynamics.
Finally, the platform is used to evaluate opposition control and several reinforcement learning
algorithms including DDPG, PPO, and DQN for real-time jet actuation aimed at reducing skin-
friction drag. While the DDPG controller correctly identifies causal relationships between actuation
and downstream wall-shear stress, none of the RL algorithms produce systematic drag reduction
within the explored training horizon, owing to long convective delays, turbulent variability, and
limited sample efficiency. These experiments nevertheless validate the end-to-end interoperability
of STREAmSML with modern RL frameworks and establish a reproducible benchmark for future
algorithmic innovation.
This work provides the first open, GPU-enabled, CTF-inspired platform capable of supporting
sensing, modeling, and control research in supersonic boundary layers and SBLI. Its modular, high-
performance design positions STREAmSML as a foundation for future advances in data-driven flow
control in compressible turbulent regimes
Localizing Visual Allodynia in Migraine
No full textA hallmark symptom of migraine is photophobia, a sensitivity and aversion to light that can occur during and in between migraine attacks. Visual stimulation can reportedly trigger migraines suggesting a link between visual processing and migraine pathophysiology. This study examined the neural correlates of visual processing in migraine. Previous studies have found that visual allodynia, pain, and discomfort in response to innocuous stimuli in migraine is associated with hyperexcitability in the visual cortex. Recently, abnormal activity in retinal receptors has been identified in migraine, at the level of the cones, rods, and retinal ganglion cells. This challenges the cortical hyperexcitability hypothesis of photophobia. We examined the cortical and retinal responses to visual stimulation in individuals with migraine to identify where along the visual pathway hyperexcitability is first expressed. We measured electroencephalography (EEG) and electroretinography (ERG) activity in 26 migraine and 16 headache-free individuals. EEG over visual cortex was used to measure cortical responses to flickering light (28.3Hz). To assess the activity from the retina, we used ISCEV standard methods for assessing cone, rod, and retinal ganglion activity. Specifically, to assess rod activity, participants were dark adapted and presented light at low luminance 0.28 Td and 85 Td. To assess cone activity, 28.3Hz and 1.96Hz bright flicker were presented to maximize cone driven responses. To assess retinal ganglion activity, we used red-blue light at 3.4Hz to identify the photopic negative response (PhNR). Stimuli intensities were modulated by measuring participant pupillary response to maintain consistent retinal stimulation. We found that individuals with migraine exhibited moderately greater cortical responses to visual flicker than headache-free individuals, consistent with the cortical hyperexcitability hypothesis. At the level of the retina, we found that individuals with migraine displayed abnormal retinal ganglion cell activity to visual flicker, implicating retinal cells in migraine-associated photophobia. This is one of the first studies to comprehensively examine visual allodynia at multiple levels of the visual pathway in the same individuals. Our results may have implications for targeted treatment for visual allodynia and migraine
From Practice to Process: Formalizing API-First Design through Method Engineering
No full textThe evolution of distributed and cloud-based systems has led the computing community to converge on Microservice Architecture (MSA) as a preferred solution to distributed software design. An expected result of microservice design is a well-defined Application Programming Interface (API) that facilitates access to microservice and system capabilities. Despite the ubiquity of MSA, the availability of software design methods supporting MSA, extensive documentation, and defined frameworks that guide practitioners in implementing microservices, challenges remain in defining concise microservice boundaries and exposing clean APIs for feature access within microservices. In response to these challenges and opportunities, API-First Design (API-FD) is emerging as a viable approach to microservice and distributed system architecture. API-First principles suggest that the capabilities and features of a system are identified early in the design cycle and exposed to external parties (e.g., end-users and third-party developers) via a set or suite of stable, well-defined, and well-documented APIs. The principles also recommend that API consideration, design, and implementation serve as the foundation and core of system design. A significant challenge associated with API-First Design is its infancy, limited peer-reviewed research, and an implied rather than well-defined understanding among industry practitioners and researchers. In this dissertation, we advance the state of API-First Design through rigorous research in the context of multi-client microservice-based systems and, in doing so, provide a research-based foundation that improves its accessibility for our academic peers and industry partners. To build this encompassing foundation, we explore and document the benefits of API-First Design, propose a formal definition for the term "API-First Design," and construct a set of supporting resources: a dictionary, a taxonomy, and a process as the pillars for the adoption of API-First Design as a viable software engineering approach to multi-client MSA and applications. Furthermore, we present our proposed PhiDo API-FD approach and illustrate its use in creating a proof-of-concept software product, a multi-client real estate transaction management solution. Together, these new artifacts and contributions to the art of API-centric design provide a foundational understanding of API-FD and a set of practical tools for both our academic peers and industry collaborators
Novel Body-Selective Regions Responsive to Bodies Away from the Center of Gaze
No full textProcessing of human bodies in naturalistic contexts is fundamental to social perception. Here, we report two previously undescribed body-selective regions in the human brain using 7T fMRI data from the Natural Scenes Dataset: the Ventromedial Body Area (VMBA) and the Medial Body Area (MBA). We localize these regions with functional contrasts and characterize their selectivity with encoding model analysis of BOLD responses to thousands of naturalistic images. Both regions are located near scene-selective regions, with VMBA near the Parahippocampal Place Area and MBA near the Retrosplenial Complex. Unlike known body-selective regions (e.g., the Extrastriate Body Area), which prefer single, foveated bodies, VMBA and MBA prefer multiple bodies away from the center of gaze. Thus, these regions are well situated to process the bodies we encounter in crowds, sports, and social gatherings. These findings advance our understanding of body representation, and suggest the existence of specialized pathways for processing bodies in different spatial contexts
Judges' Perceptions of Self-Represented Litigants: Addressing a Growing Concern for the Judiciary
No full textThe number of people accessing the courts without the help of an attorney is increasing dramatically. Recent cases (Johnson v. Sundeen, Case No. A25-0159, slip op. (Minn. App. Aug. 11, 2025) and Hamwi v. First Student Services, Case No. 126,272, slip op. at 12-13 (Kan. App. August 30, 2024)) have highlighted the challenges posed by the dramatic increase in self-represented litigants. The National Judicial College (NJC) conducted an informal poll of its alumni and student judges with the question, “How would you rate your level of comfort in dealing with self-represented litigants?” Of the 567 judges who responded, 86% stated that they were either “entirely comfortable” or “mostly comfortable” dealing with Self-Represented Litigants (hereinafter SRLs), whereas 12% said they were “not very comfortable” and approximately 2 ½% stated they were “uncomfortable.” One hundred seventy-five (175) of the respondents also provided comments, which will be analyzed in this article. Judges are the gatekeepers for SRLs’ access to justice initiatives and have a challenging job of guiding SRLs through the court process. Therefore, the “level of comfort” of judges when dealing with this issue is perhaps the most important factor in determining how effective any provided aid will be for SRLs
Epistemic Injustice Toward Women in Engineering
No full textWomen are underrepresented in engineering in the United States. To investigate this problem, engineering education research (EER) has examined possible hindrances to women pursuing an engineering degree and being retained in the field. One such hindrance is the barrier for women in the domain of knowledge and knowing. The ways engineering culture values, rejects, or interprets certain knowledges may conflict with women’s ways of knowing, which may serve to discourage or prevent women from entering or staying in the field. Epistemology—the theory of knowledge and knowing—and other epistemic theories are gaining popularity in the EER community, and researchers are looking to the epistemic dimensions of engineering education to understand the cultural foundations of engineering that may not align with policies concerning diversity, equity, inclusion, and justice. One such theory researchers are utilizing explores justice focusing on epistemic interactions, aptly named epistemic injustice. This work offers a research design to examine a possible explanation for women’s underrepresentation and oppression in the field: the devaluation of women’s knowledge uncovered by the theory of epistemic injustice. The consequences of women’s devaluation may manifest in epistemic beliefs, informing their personal epistemologies, which I investigate in this dissertation.The purpose of the research study is to tell the stories of undergraduate women engineering students, specifically their experiences of epistemic (in)justice (EIJ) and the resulting impact on their personal epistemologies (PE) in engineering contexts. The following are the research questions used to guide this dissertation:
RQ1. How do women engineering students describe their experiences of epistemic (in)justice, specifically testimonial and hermeneutical types?
RQ2. How do women engineering students interpret the impact of epistemic (in)justice on their own personal epistemologies in the context of or regarding the engineering field?
I employed narrative analysis to explore undergraduate women engineering students’ stories and produce contextually specific epistemic snapshots of engineering for women. I used three semi-structured interviews with each of 10 participants to collect narrative data about their epistemic experiences and PEs. After multiple rounds of narrative smoothing and member checking, I co-constructed with the participants 10 narratives telling stories of five participants’ experiences of EIJ and the evolution of their PEs. From comparing across narratives, I developed a conceptual model that shows the process-based relationship between the epistemic constructs described in the narratives. After mapping the conceptual model on existing models for cultural formation in engineering education, I answered a call to action to produce inclusive environments for women by fostering epistemic justice through fair knowledge-based interactions.
The documentation of what EIJ looks like for women in engineering provides a foundation in EER to understand the role EIJ and epistemology play in women’s underrepresentation and repression. Women’s stories of epistemic injustice in their engineering experiences will highlight possible misalignments between inclusive policies for women and their realities of being part of a minority population in engineering
Winter Carbon Dioxide Efflux in Sierra Nevada Meadows: Contributions to Annual Carbon Budgets, Influence of Snowpack, and Subnivean Conditions
No full textMeadows in the Sierra Nevada contain a disproportionate share of regional soil carbon and can display high levels of carbon dioxide (CO2) efflux during the growing season. Yet wintertime instantaneous CO2 efflux rates from these systems, along with further understanding of what ecological mechanisms are affecting CO2 efflux remains poorly constrained due to limited site access and hazardous winter conditions. Wintertime CO2 efflux may represent a substantial amount of the annual carbon budget because low but persistent fluxes can continue occurring beneath snowpack for several months. We quantified instantaneous wintertime CO2 efflux rates across two Sierra Nevada meadows. We measured Van Norden meadow in the winters of 2023-2024 and 2024-2025 and Martis meadow, during the winter of 2024-2025. We collected more than 300 instantaneous CO2 efflux measurements across both snow surface and subnivean environment (including soil, water, and ice). Mean winter CO2 efflux rates among both meadows, all collection dates and all surface types (snow, soil, ice and water) was ~1.5 μmol m⁻² s⁻¹. When comparing both sites in the winter of 2024-2025, CO2 efflux was highest in Van Norden at the soil surface, whereas CO2 efflux was lowest at the snow surface in Martis. CO2 efflux was similar when comparing soil surface efflux rates at Martis and snow surface efflux rates at Van Norden. CO2 efflux was significantly different between the meadows, between surfaces (snow surface and pit surfaces), and there was a significant meadow by surface type interaction. Snow depth did not significantly influence CO2 efflux at the snow surface in either meadow, whereas increasing subnivean water depth was associated with reduced rates of CO2 efflux in Van Norden meadow. Our sensitivity analysis indicated that while cumulative winter CO2 efflux estimates were sensitive to changes in instantaneous flux rates, these changes had relatively minor influence on cumulative annual CO2 efflux estimates. Together, these results demonstrate that wintertime CO2 efflux rates from Sierra Nevada meadows is substantial, influenced by site-specific and subnivean conditions, and sensitive to measurement location, emphasizing the importance of wintertime sampling approaches for improving annual carbon flux estimates
Understanding the electrification-driven lubrication mechanism of phosphonium ionic liquids to enhance the degradation resistance of electric vehicle components
No full textThe electric vehicle (EV) market has been manifolded in the recent past, and therefore, enhancing the durability and efficiency of moving mechanical assemblies (MMAs) in EVs has become increasingly important. Despite the superior efficiency of EVs over internal combustion engine vehicles, more than half of their energy loss is still attributed to friction and wear in components such as gears and bearings. Moreover, electrified environments in EVs introduce stray currents that exacerbate material loss through arcing, challenging the performance of traditional lubricants. As an alternative, conductive lubricants, such as phosphonium based room temperature ionic liquids (PRTILs) could be utilized. However, the ionic nature of PRTILs could be detrimental to the surface due to corrosion, which can accelerate the overall material loss under sliding interaction. Till date, there is barely any study where the synergistic wear-corrosion mechanisms of PRTIL lubricated surfaces have been addressed. Furthermore, the impact of stray current could deteriorate the wear-corrosion synergism, which has a significant research gap.To address these research gaps, this dissertation investigates the development and performance of aromatic ring containing PRTILs as next-generation conductive lubricants for EV MMAs. Here, at first, PRTILs are synthesized, characterized, and studied for their unique ability to reduce friction and wear under tribological interaction in a ball-on-disk setup. Further, the best performing aromatic PRTIL is compared with a popular non-aromatic halogen-based PRTIL over a wide temperature range. Next, the corrosion behavior of both aromatic PRTIL and non-aromatic PRTILs are investigated using electrochemical techniques such as open circuit potential, potentiodynamic polarization, and cyclic potentiodynamic polarization. Furthermore, synergistic material loss rates are evaluated in a combined wear-corrosion setup facilitating sliding under electrochemical corrosion. Next, the effect of external electrification on tribological behavior of the PRTIL lubricated steel surfaces are investigated for the very first time in the electrified mini traction machine. Furthermore, the impact of external electrification on wear-corrosion synergism is revealed for PRTIL lubricated surfaces using a novel in-situ test-rig. In addition to these experiments, tribological, wear–corrosion, and wear–corrosion under electrification assessments are systematically performed on steel surfaces with varying surface textures and roughness levels. All tests are carried out under non-abrasive conditions to isolate and evaluate the influence of surface texture and roughness on the coefficient of friction (COF) of PRTIL lubricants. Finally, using machine learning algorithms, the physicochemical and electrochemical properties of PRTILs are utilized to correlate with the resulting COF and synergistic material loss rates.
Key findings in the dissertation demonstrate the critical roles of molecular structure, viscosity, and wettability in enhancing the lubrication performance of PRTILs over traditional lubricants. Double ring containing PRTILs, having higher viscosity, and lower wettability were beneficial to reduce friction, and wear than single aromatic ring containing PRTILs, or halogen-based PRTILs or traditional lubricants. Structure-property correlation for PRTIL lubricated surfaces were established using a structural feature parameter that showed significant correlation with the resulted friction, wear, and surface roughness after the tribological assessment. Further, double ring PRTIL exhibited lower corrosion rates than other PRTILs or lubricants due to their better surface repassivation ability. The wear-corrosion synergistic assessment revealed that the total material loss under the combined effect of wear and corrosion, increased significantly for each PRTIL, due to the corrosion accelerated wear mechanism. The double ring PRTILs, having least corrosion rates towards surface, were able to provide significantly less overall material loss rates than other PRTILs, or lubricants. Furthermore, the surface texture and roughness assessments revealed that parallel and randomly finished surfaces exhibited lower COF compared to perpendicular and 8-ground textures. Additionally, higher surface roughness led to an increase in COF overall. To connect the physicochemical and electrochemical properties of PRTILs with wear-corrosion synergistic material loss rates, machine learning based predictive modeling framework was employed that provided dimensionless parameters using principal component analysis. The first principal component demonstrated significant impact of the material loss rates overall. It was further revealed that a greater charge transfer resistance better protected the double ring PRTIL lubricated surfaces from corrosion accelerated wear than single ring aromatic PRTIL lubricated surfaces.
Additionally, the application of external current during wear revealed that a moderate ionic conductivity is ideal to limit the wear rate for PRTIL lubricated surfaces. The ionic conductivity was also crucial in modulating electric contact resistance, where higher conductivity reduced the resistance and caused surface damage through electric shock across the wear track. On the other hand, less conductive lubricants also exhibited significantly high material loss under electrification due to arcing. Finally, the wear-corrosion synergism under electrification further supports these findings where it was revealed that moderately conductive double ring PRTIL was able to withstand the combined effect of wear and corrosion under electrification, whereas PRTILs with the highest ionic conductivity experienced significantly higher material loss rates.
Overall, this work offers foundational knowledge for designing sustainable lubricants to meet the growing demands of electrified MMAs. The outcome of this research bridges the critical gaps in understanding the interfacial mechanisms of PRTIL lubricated electrified surfaces and paves the way for future research in EV tribology and PRTIL based lubrication technologies