22689 research outputs found
Sort by
Characterizing the Role of Tip Leakage Flow in the Onset of Stall in Axial Compressors and Cavitation Inception in Ducted Propeller
Get PDFReliable design and modeling of turbomachinery require an accurate understanding of the underlying flow physics. The task becomes particularly challenging when these machines operate under extreme conditions. This thesis examines two specific phenomena – the onset of stall in axial compressors and cavitation inception in a ducted marine propeller, both of which are associated with the flow in the tip region of the machines. The experimental work utilizes performance measurements, cavitation visualization, phase-locked and high-speed stereoscopic particle image velocimetry (SPIV) in the JHU refractive-index-matched turbomachine facility, providing unobstructed optical access to the rotating blades.
In a low aspect ratio compressor, the pre-stall flow field involves the backward leakage flow extending upstream of the rotor and the radially inward expansion of the low axial momentum region throughout the rotor passage. These effects are associated with the formation of backflow vortices (BFV), which are the most prominent pre-stall flow instabilities, as shown in earlier works. Next, measurements are performed in a new high-aspect ratio, mid-chord loaded compressor rotor to study the universality of the phenomenon. The pre-stall flow is also conditionally averaged using a suitable criterion of velocity magnitude near the rotor inlet. It captures the effect of extreme transient events, which are smeared out by ensemble-averaging. Differences between the two averaging methods highlight that the flow spillage upstream of the leading edge and the formation of high circumferential velocity regions away from the blade surfaces are consistent with the occurrence and orientation of the BFVs. The Reynolds stress distribution within the rotor passage changes significantly with operating conditions. The radial stress component is the largest at higher flow rates. In contrast, the pre-stall distribution is dominated by the circumferential stress within the rotor and by the axial stress downstream of it.
In a ducted propeller, at its design and higher advance ratios, cavitation inception occurs in axially aligned secondary vortices, which are located between the blade suction side and the tip leakage vortex (TLV) and circumferentially after the trailing edge. With decreasing advance ratio, the inception shifts first to TLV and then along its core towards the leading edge. To understand the underlying mechanism, high-resolution SPIV follows the evolution of TLV, tip leakage flow, near wake, and several secondary vortices. Time-resolved SPIV at 30 kHz enables calculation of all three mean vorticity components, thus capturing the axial vortices and identifying the origin of flow structures. At higher advance ratios, inception occurs when quasi-axial vortices are stretched by the circumferentially aligned TLV and co-rotating secondary vortices located in the blade tip shear layer. With decreasing advance ratio, inception shifts to the TLV and towards the leading edge owing to earlier rollup and higher vortex strength, along with earlier breakup. This is evidenced by high turbulence in the TLV core and a decrease in its peak vorticity despite the increasing circulation
Learning Generalizable Representations for Video Classification and Retrieval
Get PDFVideo understanding has become a cornerstone of modern computer vision, with applications in robot perception, human activity analysis and multimedia search. The inherent complexity of videos—arising from diverse visual appearances, temporal dynamics, and contextual variability—poses significant challenges for building robust and practical systems. The research described here seeks to address these challenges by proposing techniques for learning generalizable representations and ensuring effective alignment across domains and modalities.
A key focus of this work is tackling the generalization challenges that arise in video understanding tasks such as action recognition and retrieval. For action recognition, this includes leveraging synthetic data and exploring how unsupervised adaptation techniques can bridge differences between training and deployment environments. For retrieval, the work emphasizes fine-grained alignment of text and video representations, enabling effective cross-modal search in complex scenarios. While these tasks differ in scope, they are united by the goal of building video understanding systems that perform reliably under varying conditions. Through these contributions, this research seeks to advance the state-of-the-art in video understanding and lay a foundation for building systems capable of handling the inherent variability and complexity of video data
THE BLOOD-BRAIN BARRIER: STRUCTURE, FUNCTION, DEVELOPMENT AND TOXICOLOGICAL CHALLENGES
Get PDFThe blood-brain barrier (BBB) is a critical component in maintaining the brain's microenvironment by selectively permitting the passage of essential nutrients and blocking harmful substances. This literature review explores the structure and function of the BBB, emphasizing its role in drug delivery challenges and chemical toxicity prediction. Key topics include the significance of electrical resistance as a measure of BBB integrity, historical research milestones, and the developmental dynamics of the BBB, particularly the formation of tight junctions and transporter expression. The review assesses various models used in toxicological studies to replicate BBB properties, the impact of barrier disruption on BBB permeability, and strategies for therapeutically opening the BBB to enhance drug delivery. Misconceptions about the BBB, current research gaps, and future directions for improved experimental models are addressed. This synthesis highlights the evolution of BBB research and its implications for neuroscience and pharmacology, offering insights into the BBB's critical role in brain health and disease
Growing Together or Growing Apart? Economic Security, Geopolitics, and the Evolution of U.S.-EU Relations
Get PDFUnder what conditions does the relationship between the United States and the European Union improve or deteriorate? In early 2025, diplomatic incidents demonstrated the fragility of U.S.-EU relations and the fluctuating nature of economic and security cooperation. Albeit dramatic, this period reflected historical patterns and alternating instances of convergence and divergence in international relations. Over the past decade, U.S. and EU commercial relations with China and Russia have underscored the security threats of economic interdependence. This thesis examines the U.S.-EU relationship through the prism of economic security and geopolitics during the first Trump and the Biden administrations. Over a period defined by geopolitical shifts and the U.S.-China strategic competition, it highlights national security dimensions of economic relations and the fusion of domestic and foreign policy. At the interplay of interpersonal dynamics between foreign leader counterparts and institutional structures, this thesis explores how varying political ideologies, leadership styles, and degrees of institutional cooperation reconfigured U.S.-EU relations beyond the traditional confines of “trans-Atlantic” affairs. It contributes to the study of how foreign policy is made by eliminating the compartmentalization of economic and security dimensions, highlighting policy actions, and providing geopolitical and geoeconomic insights for practitioner researchers, diplomats, and private sector stakeholders. Examples derive from critical sectors with dual-use applicability such as the automotive, semiconductor, and maritime shipping industries. This thesis also aspires to reach young professionals studying international relations without the preconceived biases of the Cold War period or the set assumptions of the early post-Cold War era
Advancing practice in arts in health: Artist workforce development and certification
Get PDFThis position paper addresses the critical need for artist workforce development and certification within the field of Arts in Health. Through extensive collaboration involving NOAH members and field professionals, this work presents a comprehensive framework for advancing arts, cultural, and creative practices integrated into healthcare, public health, and community life to enhance health, well-being, and social connection. The paper introduces the Arts in Health Practitioner-Certified (AIHP-C) role and establishes foundational documents including a new Scope of Practice and Core Professional Competencies specifically designed for Arts in Health practitioners. These documents provide clear guidance on professional roles, responsibilities, limitations, knowledge, and essential skills required for effective practice across healthcare, public health, and community settings. Central to this framework is the commitment to accessibility, inclusion, and equity in professional development opportunities. The paper addresses ongoing dialogue with teaching artists and creative arts therapists to clarify role differentiation across the continuum of arts, cultural, and creative practices. Through an iterative process, the authors have developed materials that advance the legitimization and recognition of artists' work while ensuring relevance to diverse sectors working to enhance health, well-being, and social connection. This paper serves as both a call to action and practical guidance for implementing artist workforce development, providing accessible and equitable pathways for professional advancement that support the continued growth and professionalization of the field of Arts in Health
Comparative Analysis of Digital Preservation Strategies: A Case Study of the Seoul Museum of History and the Visual Studies Workshop
Get PDFThis study explores how institutions of varying size and structure implement digital preservation strategies by examining the Seoul Museum of History (SMH), a government-funded national agency, and the Visual Studies Workshop (VSW), a small nonprofit focused on media art. The research addresses the central question: How do institutional size, resources, and mission influence preservation practices, especially under constraints of technological obsolescence, media fragility, and limited access frameworks? A qualitative method was employed, including case study analysis, structured interviews, and assessment frameworks such as the NDSA Levels of Preservation and the DPC RAM. While SMH benefits from formal infrastructure, standardized metadata, and national funding, VSW compensates through grassroots innovation, open-source tools, and community partnerships. Despite different capacities, both institutions face similar preservation challenges and demonstrate adaptability through digitization, metadata management, and engagement strategies. This study contributes to the digital curation field by offering insights into scalable, mission-driven solutions for institutions facing resource disparities and emphasizes the importance of flexibility, metadata, and collaborative networks in long-term digital preservation efforts
ML-Assisted Data Assimilation in Transitional High-Speed Boundary Layers
Get PDFNatural transition in high-speed boundary layers begins with the exponential amplification of infinitesimal instability waves. As these disturbance amplitudes grow, nonlinear interactions emerge, eventually leading to turbulent breakdown. Data assimilation provides a powerful framework to integrate sparse experimental measurements with Direct Numerical Simulations (DNS), enabling improved prediction of such complex, multiscale flows. In transitional high-speed boundary layers, one seeks to assimilate wall pressure measurements obtained using piezoelectric sensors (PCB) to infer the full flow field. This process requires solving an inverse problem and also demands efficient evaluation of the Navier–Stokes operator. This thesis develops a data assimilation framework for high-speed transitional flows using Deep Operator Neural Networks (DeepONet).
The first part of the thesis demonstrates how DeepONet can be used for forward modeling, predicting downstream flow in the nonlinear growth regime from upstream disturbances. Two cases are considered: a Mach 4.5 boundary layer with calorically perfect gas, and a Mach 10 flow with chemically reacting gas. The performance of DeepONet is evaluated using operator-specific metrics, and results show that for more complex flow physics, DeepONet offers greater benefit for fast, accurate prediction compared to DNS. The second part addresses the inverse problem: assimilating wall pressure measurements to estimate upstream instability characteristics. This problem is formulated as an optimization problem and solved using Bayesian Optimization (BO). An ensemble of DeepONets is used to construct a statistical surrogate model of the cost function. The reconstructed flow fields closely match reference DNS, and the method significantly outperforms conventional Kriging-based BO in computational efficiency. The final part extends the framework to predict transition location from inflow spectra using a Bayesian DeepONet. Because the transition Reynolds number is inherently stochastic, a latent-variable formulation is adopted to predict both the mean and uncertainty of the transition Reynolds number. An active learning strategy selects the most informative training samples, reducing computation cost by approximately 47\% compared to uniform sampling. By combining DNS, machine learning, and uncertainty quantification, this thesis advances data assimilation in high-speed boundary layers and provides an efficient tool for transition prediction and flow reconstruction
THE ROLE OF MULTIPLE SOURCES OF INFORMATION IN INDIVIDUAL DECISION-MAKING FOR REVERSIBLE MODERN CONTRACEPTIVE USE: A STUDY OF MASS MEDIA AND SOCIAL MEDIA IN INDIA
Get PDFIndia maintains one of the highest rates of female sterilization in the world despite reversible modern contraceptives (RMCs) being widely accessible. Evidence points to exposure to family planning information influencing women’s decision to use RMCs, but studies have thus far only considered individual sources of information in isolation. Framing the multitude of information more accurately as a complex adaptive system, this dissertation investigates the simultaneous influences of two salient sources on RMC use: mass media and social media.
A scoping review was first conducted to identify past approaches to measuring the influence of social media use (SMU) on health behavior in India. Next, Twitter was scraped for geotagged, English-language tweets posted in India that discussed family planning or RMC methods. The tweets were analyzed for sentiment using a trained ChatGPT model and then aggregated and compared across 11 states. Finally, these Twitter-derived variables were combined with National Family Health Survey data in multilevel logistic regression models to assess the associations of social media and mass media, both separately and jointly, with RMC use.
The scoping review returned only 21 studies connecting SMU and health behavior, with little evidence for causality and no consideration of family planning. The corpus of tweets analyzed was also small (N = 2,084) but revealed significant differences in the distribution of sentiment (i.e., supportive, oppositional, neutral, and indeterminable) across states. Tweets with general keywords (e.g., “family planning”) were more frequently supportive than those discussing specific contraceptive methods (e.g., “oral contraceptive pill”), perhaps reflecting the broader normative environment in India which has supported family planning for decades largely through female sterilization. Finally, the regression results indicated significant associations between RMC use and Twitter discourse sentiment (scaled AOR: 0.62, 95% CI: 0.48 – 0.81) and mass media exposure (across all categories) separately, but not together as an interaction term. The negative association between RMC use and Twitter sentiment suggests reverse causality, possibly indicating that states with lower RMC use have stronger advocacy efforts on social media.
Further research is needed to characterize the complex interactions between sources of family planning information and their influence on RMC use in India
Identifying and Understanding the Dynamical State of Clusters of Galaxies
No full textThe study of the dynamical states and mass accretion histories of clusters of galaxies touches on several key fields in astronomy. Possible tensions in the current cosmological model require a better understanding of the dynamical state to be rooted out. Several topics in astrophysics, including the evolution of galaxies in dense environments, the behavior of active galactic nuclei, gas dynamics, the cooling of cluster cores, and the properties of dark matter all require or benefit from a finer understanding of galaxy cluster mass accretion history. Previous work has sought to refine our understanding of dynamical state through improved observations and detailed simulations. In my dissertation work I have integrated a new tool, machine learning, into the study of galaxy cluster dynamical state and I have re-contextualized challenges related to observations and simulations with that tool in mind. I have demonstrated that machine learning can be a powerful tool for improving observations, by efficiently using all-sky surveys to identify potentially dynamically active clusters that would benefit from follow-up observations. The tool established by this work is one that can be fine-tuned for a variety of use cases and telescopes, including future X-ray observatories. In the subsequent study, I investigated the robustness of mass accretion histories in an N-body simulation. Simulations are key to improving our understanding of cluster formation and to the development of any mass accretion history constraining machine learning model; therefore I used the results of this work to more broadly assess the reliability of mass accretion histories in simulations. Finally, I established that galaxy cluster mass accretion rate can be constrained from observations of the intracluster medium using machine learning. This study confirmed the importance of asymmetry, density profile, and substructure in the estimation of mass accretion rates. It also discussed the possible benefits and difficulties in adapting the model to real observations. Such a program will require better observations, reliable simulations, and effective machine learning models
Oral History Interview of Laurie Zoloth
No full textThis interview with Laurie Zoloth, PhD, is part of “Moral Histories: Voices and Stories from the Founding Figures of Bioethics,” an oral history project of the Johns Hopkins Berman Institute of Bioethics. Professor Zoloth is the Margaret E. Burton Professor of Religion and Ethics at the University of Chicago. She is the author of four books and co-editor of six others, including Second Texts and Second Opinions: Essays on Jewish Bioethics and An Ethics for the Coming Storm: Jewish Thought and Global Warming. Her research explores religion and ethics, as well as the bioethics of genetic engineering, gene drives, stem cell research, synthetic biology, and climate change.
Prof. Zoloth discusses growing up in Los Angeles influenced by her post-war immigrant Jewish parents and the political activism of the 1960s. She describes going on the first Venceremos Brigade to Cuba, traveling by invitation to Maoist China as a trade unionist in the early 1970s, and becoming a licensed vocational nurse in Philadelphia. After moving to Berkeley, CA, Zoloth discusses her growing interest in philosophy, religion, and ethics at San Francisco State University. She describes her involvement in founding various organizations like the Society for Jewish Ethics and the International Society for Stem Cell Research. Zoloth recounts her experiences with “bioethics summer camp,” a relaxed event where bioethicists could discuss complex issues together. She details her work as an investigator and chair of the Howard Hughes Medical Institute’s Bioethics Advisory Board, most notably during the 9/11 attacks, and as a bioethics advisor for NASA on issues like planetary protection, the ethics of space exploration, and the “Sundowner Report” on animal ethics.
She discusses her teaching and scholarship at Northwestern University and the University of Chicago and reflects on the challenges of balancing her career while raising five children. The interview concludes with a discussion of the role of bioethics in anticipating and reflecting on the ethical implications of climate science, scientific research, and public health. In her role as a scholar of ancient religious texts, she ends the conversation with thoughts about the importance of memory and prophecy in philosophy and theology