116964 research outputs found
Sort by
The spatial neutron algorithmic model
Get PDFThe Spatial Neutron Algorithmic Model (SNAM) represents a novel numerical technique which leverages random forest regression models to improve both the precision of passive, neutron-based nondestructive assay (NDA) estimates relative to what can be achieved with the standard Point Model. This advancement is facilitated by the emergence of full list-mode, multi-channel data acquisition technologies in neutron detectors.
Traditional assay techniques use the Point Model to translate observed multiplicity count rates into SNM mass estimates. However, uncertainties regarding the spatial distribution of SNM within the detector and the material composition of the waste matrix undermine two of the key assumptions made under the Point Model, resulting in assay uncertainties up to 30%. Using full list-mode, multi-channel outputs, SNAM can examine the ratio of count rates between any and all channels/channel combinations to estimate the effective efficiency of the detection. Additionally, if represented visually, these ratios can be used by a trained individual to estimate the spatial distribution of source material within the detector.
In being able to determine the efficiency of the measurement (rather than simply assuming an often inaccurate, static value), SNAM decouples the Point Model from the limitations of its assumptions, leading to improved assay precision. Additionally, SNAM expands the capability of neutron detectors beyond just NDA measurements – the decoding of spatial information from measurements grants waste inspectors access to new data streams which can be used to better understand their system and inspect their items in novel manners.Mechanical Engineerin
Beauty, but first, joy : using reflective practice framework to explore the intersection of textile arts and inclusive education
Get PDFHow does one center process when teaching product-based skillsets like sewing? Does creating something beautiful have to come at the cost of the maker’s joy? These are questions that converged at the intersection of three roles that I’ve played in my life: a theatrical costume designer, an accessibility aide, and a teaching artist. This MFA thesis project reflects on my personal experience in creating curricula that integrate textile artistry and inclusive education and dreams of the creation of spaces that combine these practices. At the convergence of these practices, I seek to explore the question: How can I use reflective practices to develop my teaching artistry and to create inclusive textile education spaces centered on process and exploration? To investigate this question, I began by reflecting on and analyzing my creation and implementation of a sewing workshop at an adult arts education program. Building on my findings from this workshop, the project culminated in the creation of a prototype communal education space for textile arts practice. In sharing and reflecting on these experiences, I hope to improve my efficacy as an inclusive educator and artist and to create resources for others with similar goals.Theatre and Danc
Whose garden? : experience and identity in the House of Venus in the Shell (II.3.3), Pompeii
Get PDFThis paper reexamines the house at II.3.3 in Pompeii, departing from previous scholarship by framing the conditions of lived experiences after the earthquake of 62 CE. Chapter 1 situates the homeowners—the Lucretii Valentes—within Pompeii's social fabric, highlighting the role of resilience in post-earthquake construction and redecoration. The family's social and economic connections, evidenced by their involvement in priesthoods and hosting of post-disaster games, underscore their influence. In Chapter 2, the focus shifts to the domus' role in identity formation, which I approach by examining Valens' decor choices. His selection of imagery, notably Venus, reflects cultural influences and elite aspirations. The significance of gardens in Roman society is then explored, revealing their multifaceted roles and symbolic meanings. Chapter 3 delves into daily activities within II.3.3 through viewing scenarios, considering diverse household perspectives, including an enslaved woman's. By incorporating these viewpoints, the study unveils the potential of the decor to evoke multiple interpretations and experiences. Overall, the paper constructs a narrative of resilience and identity for the inhabitants of II.3.3, exploring the intricate interplay between art, identity, and resilience in the ancient city.Art Histor
"It feels like having PTSD" : how race-related stress and subjective school racial composition are impacting Black college students' mental health and well-being
Get PDFRace-related stress is an unfortunate yet pernicious force in the lives of Black people in America (Jones et al., 2020). As anti-Blackness has been systemically internalized within the fabric of America, one institution where prior research has examined the manifestation of anti-Black rhetoric is the American education system, specifically in post-secondary education (Strauss et al., 2022). This systemic racism is often manifested most commonly through intergroup/interpersonal experiences of race-related stress from non-Black faculty, staff, and peers and through racist institutional practices from the college/university itself (Bourke, 2010). Finally, while this is less commonly addressed, we also see internalized racism manifest through intragroup marginalization on these campuses (Castillo et al., 2007; Smedley et al., 1993), such that Black young adults may internalize and further perpetuate anti-Black rhetoric. Considering the current sociopolitical climate, multiple recent racial reckonings, internalized racism, and the new recurring threat of the eradication of diversity, equity, and inclusion practices on various college campuses, Black college students are at the epicenter of it all.
Further, racism has known deleterious impacts on Black students’ mental health (defined by this study as self-rated depressive symptoms and anxiety) and positive well-being (defined by this study as self-rated flourishing and life satisfaction), (Paradies, 2006). However, to better understand the differential health implications of different race-related stress experiences on Black students’ mental health and positive well-being, it is crucial to recognize how social context, namely racial composition, is associated with the relationship between race-related stress and health outcomes. Prior research has suggested that perceiving being in varying proximity to those who share similar and different racial and cultural experiences can influence Black people’s mental health and positive well-being and the kinds of race-related experiences they have (Boykin et al., 1986). Therefore, it is essential to highlight how perceived or subjective racial composition is associated with the relationship between race-related stress and these Black students’ mental health and positive well-being.
This work is rooted primarily in A.W. Boykin’s Triple Quandary Theory (1986), Lipsitz's (2007) notion of “racialized space,” and Gloria and Rodriguez’s (2000) Psychosociocultural (PSC) framework. I build upon these frameworks by identifying how being surrounded by those of similar or varying sociocultural experiences can help, hinder, or exacerbate the effects that race-related stress has on Black college students’ mental health and positive well-being. Specifically, in a US-based sample of 428 Black college students, I tested the direct associations between institutional, intergroup, and intragroup race-related stress experiences on Black college students’ mental health and positive well-being. Next, I tested whether there were patterns in how students perceive campus racial composition (i.e., latent profile analysis) and whether this subjective perception directly influences Black students’ positive well-being and mental health. Further, I examined how the relationships between race-related stress experiences and mental health and well-being differed according to profiles of students' subjective school racial composition. Finally, I also explored how these relationships differed according to respondent gender, as gender is also associated with Black people's race-related stress experiences (Crenshaw, 1991). I found that institutional race-related stress experiences were associated with higher depressive and anxiety symptoms, as well as a lower sense of flourishing. While correlation analyses yielded opposite findings, direct effects showed that intergroup race-related stress experiences were associated with higher reported sense of flourishing among Black students in the sample. Intragroup experiences of race-related stress were associated with higher depressive symptoms and a lower sense of flourishing. I also discovered three unique subjective racial composition profiles (i.e., More Black, More White, and Diverse), and these profiles were significantly associated with student mental health and well-being such that the Diverse profile saw the lowest rates of flourishing and highest rates of depressive symptoms and anxiety compared to the other two profiles. However, Black students in the Mostly White profile experienced the lowest levels of life satisfaction compared to the other two profiles, while the Mostly Black profile saw the highest levels of life satisfaction. There were marginally significant differences in the relationship between institutional race-related stress experiences and satisfaction with life according to profile membership, and there were no differences between race-related stress experiences, mental health, and positive well-being according to student gender identity.
Identifying and understanding the roles of these factors in Black college students’ lives is extremely important, as mental health and positive well-being are crucial indicators of their success and longevity in their academic pursuits and everyday lives (Cokely et al., 2023).Human Development and Family Science
High speed film cooling parameters and optimization
Get PDFIn modern gas turbine engines, hot gas Mach number Ma [subscript ∞] in the turbine section may be transonic. However, almost all film cooling studies to date have been performed using low-speed flows, often with Ma [subscript ∞] < 0.1. This reasoning was valid, as past works suggested film cooling performance was not dependent upon Ma [subscript ∞], notably when using cylindrical film cooling holes. However, recent computational and experimental evidence with shaped film cooling holes would suggest that in-hole flow structures that occur at high Ma [subscript ∞] can degrade film effectiveness. At engine realistic Ma [subscript ∞], the flow within the hole can be supersonic, and accompanying shocks and separation features not seen in low-speed flows may be present. In light of this, a new high-speed wind tunnel facility was recently constructed which permits film cooling experiments with Ma [subscript ∞] up to 0.58. Using this facility, film cooling performance of a 7-7-7 shaped film cooling hole was evaluated relative to varying mainstream Mach number, from Ma [subscript ∞] = 0.15 to Ma [subscript ∞] = 0.58. Film cooling performance was also evaluated relative to varying stagnation temperature ratio, or density ratio by analogy. In addition, the implications of scaling with blowing ratio and pressure ratio are discussed. Computationally, adjoint based optimization was used to develop novel optimized film cooling geometry for both high-speed (Ma [subscript ∞] = 0.75) and low-speed (Ma [subscript ∞] = 0.15) conditions. Although some geometric features were shared between the low-speed and high-speed optimizations, such as external wall protrusions, each geometry was distinct. The high-speed optimization had smaller, asymmetric protrusions, and a smaller hole area than the low-speed optimization. Experimentally, the high-speed optimization outperformed both the 7-7-7 shaped hole and the low-speed optimization across all Ma [subscript ∞] tested.Mechanical Engineerin
A hardware-LabVIEW integration framework for real-time measurement in scaled power grids
Get PDFThe design, implementation, and validation of a three-phase measurement system using voltage/current sensors, a resistor board, NI-9220 data acquisition modules, and a grid simulator are presented. The primary objectives were to downscale high-voltage signals input (up to 225 V line-to-line, peak) and high-current signals to levels suitable for data acquisition,
and to achieve precise frequency synchronization via a Phase-Locked Loop (PLL). LabVIEW and PLECS software were employed to develop a PLL-based control strategy for real-time data capture and frequency synchronization at 60 Hz. Experimental results confirm that voltage and current signals are reliably scaled below the NI-9220’s ±10 V limit, and that the PLL maintains stable operation with minimal frequency deviation.
To ensure accurate frequency tracking, an amplitude-invariant Clarke/Park transform feeds a PI-based PLL, whose gains were first tuned in PLECS. A crossover frequency of 30 Hz led to Kp = 18.84 and Ti = 0.0919. Simulation results showed that these gains keep the PLL frequency within ±0.5 Hz (about 0.52% deviation) of the nominal 60 Hz. LabVIEW experiments validated the design for a 100 V, yielding P ≈ 224.6 W and Q ≈ −16 VAR.
This integrated experimental hardware and simulation software setup validates the accuracy of the hardware configuration, aligns with theoretical derivations, and provides a robust and flexible framework for real-time measurements in a scaled power grid.Electrical and Computer Engineerin
Interpolated second order relative motion models for arbitrary spacecraft trajectories
Get PDFRelative motion dynamics is the discipline of computing the trajectory of one spacecraft relative to another. Since the early days of the space age, successful application of relative motion has been a fundamental operational capability for space programs. Relative motion is intrinsically useful for proximity operations, but it also has great utility in enabling the solution of other more general problems in estimation, control, optimization, and more. However, the changing landscape of space missions places increased stress on standard relative motion models which are typically either fast with limited accuracy or slow with exquisite accuracy. In addition, these models are often restricted to a narrow class of reference trajectories or dynamical environments. One way of circumventing the speed/accuracy dichotomy is by adding memory footprint as a new design dimension. In this work, a relative motion framework that is fast and accurate is developed through the use of continuous interpolation of partial derivatives of a reference trajectory flow. These flow derivatives are also called state transition tensors (STTs). Terms through second order are computed and retained. The motion model is also general: it is applicable to arbitrary reference trajectories and a large class of relative force models. The resulting system is especially useful for exactly periodic orbits in which case it can provide relative motion solutions for all time. In the case of an arbitrary non-periodic trajectory solutions are available between two arbitrary times within a prespecified problem time domain. First, the viability of an interpolated STT model is investigated by generating a continuous second-order relative motion model for a simplified Gateway near-rectilinear halo orbit (NRHO) in the context of the circular restricted three-body problem (CR3BP). The techniques for generating and storing a minimum set of STTs for the model are verified, along with the procedure for constructing STTs between arbitrary times in the orbit from the minimum set. The utility and impact of regularization on quadratic relative motion modeling is also examined. The resulting model is called QIST, for Quadratic Interpolated State Transition. The reference trajectory and solutions to the variational equations are stored in memory, enabling rapid calls to retrieve relative motion around the reference without further use of a numerical integrator. Versions of QIST using two interpolation methods are examined: Chebyshev polynomials and dense Runge-Kutta splines. The QIST models are compared against a legacy first-order numerically integrated state of practice model in terms of speed, accuracy, and memory footprint. The spline version of QIST works directly in the time domain and requires 6.5 MB to achieve 14 digits of interpolant accuracy, while the Chebyshev version requires regularization but only consumes 130 kB of memory for the same result. At runtime, the QIST models demonstrate speed improvements over state of practice methods requiring numerical integration. The Chebyshev QIST model is faster than the state of practice for integration times longer than 0.2 Gateway periods, while the dense spline QIST model surpasses the state of practice model regardless of time of flight or integration accuracy settings, by up to two orders of magnitude. Solutions propagated to first order with QIST agree with the standard of practice, while second order solutions show order of magnitude accuracy improvements throughout almost all of the Gateway orbit. Next, the QIST model is generalized to use a SPICE SPK kernel as the reference trajectory. The SPICE toolkit is a powerful and widely used system for storing, visualizing, and computing the geometry of spacecraft and celestial bodies published by JPL. The QIST dynamics model is also expanded to incorporate ephemeris gravity as well as spherical harmonics models for the central body. Both the linear and quadratic models are based on the precomputed SPK reference. Chebyshev approximation is discarded, so the model is generated using a dense integrator that provides a continuous interpolation between steps with the same accuracy as the integration. The SPICE referenced ephemeris relative motion model is validated against known relative motion solutions including Keplerian and perturbed motion. The elements of the kernel referenced flow partials are shown to retain 8 digits of accuracy in the worst case tested. The output of QIST is also shown to agree with standard Keplerian relative motion models. Subsequently, a Lunar Gateway relative motion model is computed for a mission-derived SPICE trajectory where the originating model is not exactly available. Another application of the new relative motion framework is demonstrated using the generated Gateway motion model with the nonlinear propagation of covariance around the Gateway orbit. Finally, an exhibition of QIST in outer planet systems is performed using the Cassini mission kernel, propagating non-Gaussian trajectory distributions through the historical Titan encounter 3. The kernel-referenced relative motion model is shown to provide accuracy improvement over non-kernel-referenced equivalent models. Additionally, significant speed improvements are demonstrated when comparing QIST to numerically integrated models. Finally, as a demonstration of the capabilities of kernel-referenced relative motion, QIST is applied to an operationally relevant mission design problem. Robust optimization is performed on trajectory correction maneuvers during the lunar lander return phase of an Artemis mission, treating the trajectory from two hours after low lunar orbit departure to arrival in the vicinity of the Lunar Gateway as a relative motion problem. The underlying dynamics propagation is performed by QIST, enabling rapid stochastic optimization techniques requiring many candidate trajectories. Multiobjective optimization is enabled by using linear covariance (LinCov) software in a simplified operational context, taking into account the availability of navigation sensors with varying measurement models, ranges, and accuracies. No numerical integration is used, since the relative motion around Gateway is fully characterized with the a priori computation of the QIST coefficients. The speed afforded by QIST enables the optimization problem to be solved via a grid search of the space of feasible solutions. Maneuver placements are computed to optimize the minimum 3σ delta-v of the trajectory, the position dispersion at a target point, and a Pareto optimal combination of these two metrics. The particular optimization problem considered in the final section was previously not attempted due to the time necessary for propagating the necessary trajectories. Combining LinCov with QIST provides dramatic speedups, enabling the computation of the optimal solution. All QIST results are shown to be in-family with legacy methods. The tradespace for optimal delta-v design is found to range from 77.0 to 93.9 m/s, while the range of optimal target point dispersion is between 1.4 and 12.2 km. The source code of QIST is made available in an online repository.Aerospace Engineerin
The mental health of Latines in college
No full textThis paper is a systematic review of the literature regarding research on Latine college students and their mental health. The review covers prior research on the topics of discrimination and acculturation, familismo and other cultural values, disparities across genders and immigration statuses, and treatment seeking and coping. Then, lessons learned from the COVID-19 pandemic are included, as well as the implications of Latine mental health for institutions of higher education. The review concludes with a discussion on the limitations of prior research and suggests new directions using the stress process paradigm.Sociolog
Terrain adaptation for autonomous navigation and model predictive control
Get PDFDeriving analytical dynamics models for real-world autonomous driving is challenging, especially for off-road navigation where terrains can be rocky, uneven, and slippery. One alternative is to learn a dynamics model for each terrain using data. However, these models only train on a fixed set of environments, and, as a result, fail when exposed to new dynamics. In real-world operations, dynamics can change frequently and unexpectedly, so robots must be able to adapt in real time. This work proposes real time terrain adaptation using function encoders, where an encoder learns to span a space of dynamics offline using data from different terrains. Then, when exposed to a new environment at run-time, the encoder infers a dynamics model using its learned representation. We demonstrate terrain adaptation in a Unity-based robotics simulator, where a vehicle must navigate across terrains with different friction coefficients to reach waypoint goals. A function encoder learns a representation of the varying terrains, and then a model predictive path integral (MPPI) control algorithm drives the vehicle, using the trained encoder to calculate trajectory rollouts. Initial results evaluate how well the function encoder adapts to new terrains based on MPPI's waypoint navigation performance in Unity.Aerospace Engineerin
Inter-operable bi-directional fractional power conversion system for battery-to-battery charging applications
Get PDFWith the increase in demand for electric vehicles, the need for innovative battery charging solutions is expected to increase. In this thesis, we explore one such unconventional charging solution that facilitates power transfer between two 400 V or two 800 V Li-ion batteries (the two common vehicle powertrain architectures). The unit will be able to transfer power between batteries in both directions regardless of which voltage is higher (or if they are equal). The unit will use “fractional power processing”, which would replace the need for a fully rated power converter with a smaller power rated power converter.Electrical and Computer Engineerin