116964 research outputs found
Sort by
Bombs, Bots, and the Principle of Distinction: The Law of Armed Conflict and Contemporary Warfare
Full text linkCritics of autonomous weapons systems (AWS) argue that these weapons cannot reliably distinguish between legitimate targets and those protected from attack. As a result, the use of AWS seems to violate the principle of distinction under international humanitarian law (IHL), which requires that combatants “not make civilians the object of attack” and not carry out attacks that are “indiscriminate in nature.” This criticism, however, misunderstands the principle of distinction and ignores important aspects of how AWS are being developed and deployed. Critics rely on an overly broad definition of AWS, and hold these systems to a standard that is inconsistent with the principle of distinction as it is actually formulated under IHL. Despite their very real limitations, the characteristics of modern AWS in fact highlight an impressive feat of technological design, and mark a further step on our long and fitful road to making warfare a less brutal and bloody enterprise.LBJ School of Public Affair
Engineering polymers for advanced technologies
Full text linkPolymer science began in 1920’s with the work of Staudinger, coworkers and others who convinced the world of the existence of macromolecules and their value. Since that beginning, polymers have played an ever-increasing role in the advancement of society. Polymeric materials have had an important role in commodity applications and in high technology. These applications invariably require the design and optimization of new polymer materials and processes. The work in this dissertation focuses on the design of polymers for three important applications: characterization of oil reservoirs, enhanced oil recovery and efficient production of microelectronic devices. Magnetic nanoparticles are expected to have a major impact in subsurface crude oil exploration as contrast enhancing agents. However, the harsh subsurface conditions often lead to nanoparticle aggregation and retention in porous media. Polyelectrolytes are often used as stabilizers to nanoparticles against colloidal aggregation in such conditions. In chapters 2 and 3, the mobility of polyelectrolyte functionalized nanoparticles has been investigated. The effect of size and polymer architecture on transport was studied. The growing demand for oil has necessitated the need for advanced extraction techniques. Polymers are often used as viscosity modifiers in oil recovery to stabilize waterflood front and increase recovery. Incumbent synthetic polymers are however limited to low salinity and low temperature sandstone reservoirs owing to their weak polyelectrolyte nature. In chapter 4, synthesis and solution properties of high molecular weight non-ionic polyethers have been explored for flooding applications. Collapse of high aspect ratio (HAR) nanostructures (found in NAND devices, Capacitors and transistors) has been a high value problem in semiconductor industry. During fabrication steps such as cleaning, liquid environments are often encountered which result in surface tension forces, pulling them towards each other once the structural/material strength of the structures is overcome. In chapter 5, stimuli responsive depolymerizable materials are explored as bracing materials to prevent collapse of HAR structures.Chemical Engineerin
Structural behavior of reinforced concrete members with cold joints : characterization and modeling in disturbed regions
Full text linkThis dissertation investigates the structural behavior of reinforced concrete (RC) members with cold joints through a combination of analytical and experimental studies. The research comprises material-level and structural-level experiments, nonlinear finite element analyses, database evaluation, and development of STM-based analytical model and design recommendation. The major achievements of this work include identifying and quantifying influential parameters governing cold joint behavior, developing design recommendations for RC structures with cold joints in disturbed regions (D-regions), and formulating a mechanics-based constitutive model for cold joints suitable for implementation in advanced numerical analysis.
The study began with an extensive literature review and an industrial survey to identify critical variables and common design practices. An experimental program was subsequently developed, including both material-level and structural-level experiments. 54 slant shear specimens were tested to quantify the effects of six key parameters on cold joint behavior. Based on these findings, two full-scale experimental programs were conducted to evaluate cold joint behavior in two- and three-dimensional structural elements, including ten deep beams and one drilled-shaft footing cap. These experiments provided detailed information about the shear capacity, stiffness, stress distribution, and failure mechanisms.
Building on the experimental results, a mechanics-based numerical model for cold joints was developed, grounded in principles of equilibrium, constitutive relationships and compatibility of interface slip. The model was integrated into a nonlinear finite element analysis software, and its accuracy was validated against experimental observations.
According to the experimental and analytical results, a load-transfer model for cold joints in D-regions was developed by incorporating the interface shear mechanics into the Strut-and-Tie Method, effectively capturing the influence of critical design parameters on cold joint behavior. The proposed model was ultimately refined into a step-by-step design recommendation, providing a practical and reliable framework for evaluating RC structures in D-regions with cold joints.Civil, Architectural, and Environmental Engineerin
Multi-threaded attracting manifold adaptive control for aerospace systems
Full text linkAdaptive control has been used to provide closed-loop tracking guarantees for dynamic systems with unknown parameters. In such cases, estimates of unknown parameters are adapted over time and these estimates are used within a judiciously prescribed control law. As part of the adaptation, adaptation gains can be increased to accelerate the rate of convergence of these estimates to their true values. In practice, however, it is not prudent to simply increase learning gains as this can have the adverse effect of amplifying any unmodeled dynamics in a true system. As such, it has been the goal of adaptive control practitioners to design adaptive control structures which provide better performance for equivalent control gains. The Multi-Thread Attracting Manifold (MTAM) adaptive control methodology developed in this work seeks to provide an improved adaptive control framework through multiple design features. The first feature is the utilization of multiple threads. This allows for the unknown parameter space to be robustly sampled, rather than relying on a single estimate. Secondly, the adaptation of each individual thread is designed with an attracting manifold design which provides desirable "no-regret" learning. These features combine to provide an adaptive controller which outperforms existing single thread adaptive controllers with identical learning gains. In this work, the baseline methodology behind MTAM is presented first for classical adaptive control and parameter identification scenarios. Secondly, forms of MTAM which accommodate uncertainties attached to the control input in both direct adaptive control and indirect adaptive control cases are presented and shown to retain performance benefits through numerical simulation. Finally, a methodology for adding and removing threads is presented which decreases the marginal computational cost of the MTAM algorithm while still providing performance benefits over existing single thread architectures.Aerospace Engineerin
Race to equity : leadership, politics, and school disciplinary policies
Full text linkThis research addresses sociological questions aimed at enhancing our understanding of the sociopolitical contexts among macro-, meso-, and micro-levels that shape high school principals school discipline policies in the racial neoliberal state. This study is informed by three bodies of theory to guide research questions and the analysis of the data. Embedded with a concept of racial neoliberalism and its connection to discipline and punishment, as the foundation of the context of this research, this research uses Foucault’s (1977) notion of docile bodies as a fruitful theoretical lens for grounding the discussion of how institutionalized power works through punishment in shaping school discipline practices, and how high school principals negotiate their disciplinary practices under the pressure of the current accountability education system in an era of neoliberal market-based reforms. The structure of this research extends the ideas of sensemaking theory to provide a novel view on the interactions of various elements of institutionalized political constraints that shape high school principals’ disciplinary practices. Using convergent mixed methods in a single case study, I investigate the following: (a) how principals make sense of pressures, constraints, and incentives that shape disciplinary practices; (b) how principals negotiate competing pressures and expectations as they implement disciplinary practices; and (c) how principals’ sociocultural identities, principals’ belief systems and sociocultural contexts inform variations in both perceptions and implementation behaviors. The findings reveal that gentrification and immigration policies bring great concerns about students’ enrollment trends and students’ social-emotional well-being. Students’ enrollment trends not only shape demographic compositions in schools but also lead to different patterns of the use of discretionary disciplinary policies. Other factors, such as statewide disciplinary policies, accountability systems, and school finance policy also put pressures and constraints on high schools’ disciplinary practices. High school principals in Austin have experienced especially strong financial constraints and use different strategies to practice the most current alternative disciplinary practices with limited budgets, yet some strategies reinforced racial biases in the school system. Findings also show that regardless of the diverse beliefs about students’ behaviors, sociopolitical views, or religion, as long as school leaders have an awareness of racial biases in the school disciplinary system and have campus administrators and teachers support to changes in disciplinary practices, schools can effectively decrease unequal disciplinary decisions to students of color. This analysis offers several directions for educational leadership practices, education policy, and research.Educational Leadership and Polic
Exploring a language-modeling approach to jargon in genre-specific instructional texts
Full text linkThe use of jargon is far-reaching not only in technical domains such as medical or engineering, but also in non-technical ones targeting a broad audience. This work introduces a dataset and an investigation of jargon in 1,014 instructions for the Free Application for Federal Student Aid (FAFSA) across 341 post-secondary institutions in the US. These instructions are crucial for students in need of financial aid for higher education, thus require both precision and low barrier of entry. Using large scale pre-trained language models, we confirm the common-sense hypothesis that jargon tends tend to be less predictable than non-jargon in language models.Linguistic
Nickel oxide nanocrystals and vanadium oxide composites for electrochromic smart windows
Full text linkThis report investigates the synthesis, characterization, and electrochromic performance of nickel oxide (NiO) nanocrystals (NCs) and vanadium oxide (V₂O₅) composites for smart window applications. NiO NCs were synthesized through a colloidal nanocrystal process while V₂O₅ films were prepared via a simple oxidative deposition from a vanadium chloride precursor. The structural analyses revealed 4 – 7 nm flowery clusters of NiO NCs and dense amorphous V₂O₅ films, each contributing uniquely to the composite's electrochromic properties. In-situ spectroelectrochemical measurements demonstrated a maximum transmittance modulation of 55% for NiO NC films, for films tested from ~400 nm thick to ~1.2 μm thick, which highlights the limitations to our fabrication technique. We present NiO films overlayed onto amorphous V₂O₅ which exhibited a two-stage oxidation process, attributed to the sequential charging of V₂O₅ and NiO layers, offering enhanced control over optical properties without requiring additional electrodes. These studies underscore the potential of NiO/V₂O₅ composites for smart windows.Chemical Engineerin
Socio-technical systems engineering and design : a meso-level network-based approach
Full text linkDifferent from traditional engineering systems design, which is keen on the design and optimization of technical artifacts, the design of socio-technical systems (STS) is guided by fundamentally understanding the complex interactions between social and technical aspects. This has posed significant challenges when applying existing systems engineering (SE) and design approaches to STS. For example, classical top-down design methodologies, such as the Waterfall model and the SE Vee model, are not appropriate for the engineering and design of large-scale STS with spontaneous interactions among individual entities or components. Although existing bottom-up design approaches are adaptable to the system scale, they primarily focus on understanding the behaviors and interactions between individual entities at the micro-level and their impact on system performance at the macro-level. In my dissertation research, the central hypothesis is that subsystems at the meso level (e.g., small clusters of individual entities) serve as critical links in system structures and could influence both macro-level performance and micro-level interactions, and thus deserve scientific investigation in STS engineering and design. However, there is a knowledge gap in understanding what meaningful subsystem information at the meso-level is and how it can be extracted and used to guide the design of an STS system to achieve the desired system performance. To fill this gap, my research objective is to develop a novel meso-level network-based framework for STS engineering and design. This dissertation is driven by answering three research questions: 1) RQ1 : How can significant meso-level system structures be identified? 2) RQ2 : What are the influences of the significant meso-level subsystems on the system performance at the macro level and the interaction mechanism at the micro level? 3) RQ3 : How can meso-level structural information be used to design an STS to achieve desired macro-level performance and micro-level functionality? The methodologies proposed to address these questions are validated through two case studies: shared mobility systems and customer-product market systems. For shared mobility systems, a network motif-based robust design framework is proposed to improve the robustness and resilience of socio-technical systems against seasonal effects. Within this framework, trip motif mining addresses RQ1, while trip motif-based system robustness metrics tackle RQ2. Formulating and solving optimization problems serves to address RQ3. Additionally, a graph neural network-based (GNN-based) link prediction (LP) model is introduced to support STS design decision-making and validation. The GNN-based model leverages local network information to enhance prediction accuracy, addressing RQ2, while implementing the LP model for design strategy validation contributes to addressing RQ3. In the context of customer-product market systems, a socio-technical system data collection framework integrating information retrieval and survey design methods is proposed to tackle the data scarcity issue in STSs. Furthermore, a novel micro-level entity design framework of STS, considering meso-level dependencies, is proposed, marking the first attempt to solve the inverse problem. This framework contributes to addressing RQ1, RQ2, and RQ3 by incorporating network motif mining, quantification of subsystem-based individual entity functionality, and entity optimization design within a unified framework. Lastly, a preliminary exploration of meso-level temporal network motifs in STS is conducted, encompassing solutions to the dynamic data scarcity issue, dynamic network modeling, and significant temporal subnetwork mining and empirical interpretation. This exploration contributes to answering RQ1 and RQ2 when considering the time dimension. Regarding the key findings and conclusions of this dissertation, we first show the effectiveness of combining information retrieval and survey design to tackle the data accessibility challenge in STS network data. Additionally, our survey study, for the first time, gathered customers' social network data alongside their purchase decision-making data, aiding in the examination of social factors influencing customers' decision-making. Moreover, while survey studies are time-consuming, leveraging named entity recognition (NER) models for mining online text data offers a viable alternative for supporting entity relationship data collection. Then, when working on the shared mobility system case study, we find that: 1) An STS's seasonal sensitivity is closely tied to imbalanced capacity planning within its subsystems. Therefore, balancing the capacity of meso-level service systems is beneficial to enhancing STS robustness against seasonal demand fluctuations; 2) The outperformance of the GNN-based predictive model, which incorporates local network information, compared to a simple neural network model lacking such consideration, demonstrates the importance of local network information in demand prediction between stations in shared mobility networks. Moreover, this outperformance persists even when network structures and density change significantly. Next, in the study of design for customer-product systems, the inter-brand triadic competition closure competition, where three products from different brands form a closed triangle competition, emerges as a significant pattern in the vacuum cleaner market system. Identifying these meso-level patterns offers a means to quantify product competitiveness. Integrating this information with network predictive models and metaheuristic approaches, like the genetic algorithm, facilitates the inclusion of local competition data in the product design process. In the study of STS dynamic analysis, we demonstrate that increasing undersampling ratios improves predictive performance, particularly in moderately imbalanced systems, enhancing the GNN-based LP model. However, in extremely imbalanced systems, a tuning process is necessary to balance computational efficiency and model performance, with the threshold-based postprocessing method consistently outperforming the rank-based method. Additionally, six temporal competition motifs are interpreted, aiding in tracking market system dynamics. In summary, my dissertation contributes to the systems science literature by introducing a novel meso-level network-based framework for STS engineering and design, thus addressing the knowledge gap pertaining to the identification and interpretation of statistically significant subsystem structures (i.e., meso-level structures formed within a complex system) and the use of such structures for STS engineering and design. The findings presented herein shed light on the importance of treating significant subsystems as crucial functional units and building blocks of STSs and underscore the need to consider them in both macro-level system design and micro-level individual entity design for optimizing system performance and entity functionality. Beyond enriching systems science from the meso-level subsystem perspective, this dissertation is expected to generate broader impacts in: 1) Addressing imbalanced source allocations in societal infrastructure systems, such as uneven distribution of public resources in urban areas. By treating local communities as meso-level subsystems and utilizing their information, this research offers policymakers actionable insights for more efficient resource distribution; 2) showing the potential to inform robust design strategies for large networked physical systems like power grids and transportation networks, the meso-level subsystem-based approach facilitates the identification of critical functional units within these systems. Subsequently, system optimization design can be guided by preserving the functionality of these identified subsystems. 3) enhancing interdisciplinary collaboration between engineering and social sciences. The frameworks proposed in this dissertation are extensible to incorporate societal analytical models. For example, in the case study of customer-product market systems, a more advanced network model that integrates customer social networks into the proposed product competition network can be easily generated to support a more in-depth analysis. By bridging the gap between technical systems engineering and social aspects, it fosters a holistic approach to addressing complex societal challenges.Mechanical Engineerin
Surfactant enhanced oil recovery in high temperature high pressure sandstone reservoir with mobility control
Full text linkSurfactant is one of the various important chemicals used for enhanced oil recovery (EOR) methods. Surfactant can be injected into oil reservoirs to reduce the interfacial tension (IFT) between the aqueous and oleic phases or to alter the rock's wettability in a favorable way, both of which increase displacement efficiency. However, in many cases, surfactant is injected with other components to control the mobility of the aqueous phase, such as polymer in SP flooding or gas in LTG flooding. Even in surfactant EOR methods, mobility control is crucial for ensuring effective sweep efficiency and oil recovery.
This thesis presents a comparative study of Surfactant-Polymer (SP) flooding and Low-Tension Gas (LTG) flooding, two advanced surfactant EOR techniques for high-temperature sandstone reservoirs. Both methods rely on ultra-low IFT to improve oil recovery, but their effectiveness also hinges on mobility control, formulation optimization, and reservoir conditions. The study also examined the optimum salinity for both surrogate and live oil, proposing a mass fraction based mixing rule to match the equivalent alkane carbon number (EACN) and optimize the salinity for microemulsion phase behavior.
For SP flooding, the research optimized microemulsion phase behavior under harsh conditions, including 100°C and high salinity (582 ppm divalent cations), using anionic surfactant formulations like TDA carboxylate and C20-24 IOS with a co-solvent. LTG flooding, involving the co-injection of surfactant and methane, generated strong foams that provided superior sweep efficiency, particularly in intermediate-permeability regions (70 mD), where SP may experience permeability reduction and reduced oil recovery.
Both SP and LTG methods achieved over 90% recovery of the original oil in place (OOIP) with live oil under well-optimized conditions. LTG requires larger chemical slugs and more time than SP but offers better mobility control in low-permeability reservoirs. SP, however, is faster and more environmentally favorable due to lower surfactant retention. The choice between SP and LTG depends on factors like permeability, operation time, and environmental impact. While SP is more time-efficient, LTG excels in low-permeability conditions, with similar overall chemical costs but higher facility expenses for LTG due to gas injection.Petroleum and Geosystems Engineerin
Energy efficient and area efficient integrated circuits design for implantable devices
Full text linkImplantable medical devices (IMDs) play crucial roles in various clinical applications. They facilitate diagnosis, deliver therapeutic interventions, and promote regenerative processes. Furthermore, IMDs are revolutionizing scientific approaches and methodologies, particularly in neuroscience research. Cutting-edge IMDs designed for neural interfaces should be able to support vital signal monitoring and advanced neuromodulation. Additionally, it is important to implement a wirelessly powered IMD with a data transmission link to eliminate the need for tethers and batteries, thus overcoming limitations in clinical use. Despite all these they can do, the IMDs are still facing several challenges, such as device miniaturization, power reduction, and closing the loop. In this dissertation, several application-specific integrated circuits (ASICs) have been developed for next-generation IMDs. In addition, in these ASICs, several circuit-level and system-level techniques have been proposed to address the aforementioned challenges. In the first ASIC design, a wireless opto-electro neural interface is developed to support simultaneous optical stimulation and neural recording. In addition, a novel voltage-boosting switched-capacitor-based stimulation (VB-SCS) and a continuous-time discrete-time (CT-DT) delta-sigma modulator-based (ΔΣM) neural recording front-end is proposed. The second ASIC design includes a novel linear-charging SCS (LC-SCS) and a CT ΔΣM neural recording front-end, for the first time, enabling a dual-modal miniaturized neural interface device. The third design presents a wireless, multi-modal physiological monitoring ASIC for animal health monitoring injectable devices. The ASIC includes an electrocardiogram (ECG), photoplethysmography (PPG), and body temperature sensing. This dissertation also explored a novel method to power the device efficiently and wirelessly. In the fourth ASIC, an ultrasound energy harvesting circuit was proposed. The ASIC is able to harvest acoustic power from a pre-charged capacitive micromachined ultrasonic transducers (CMUT) and provides the CMUT with a 4-44V bias voltage to dynamically optimize its efficiency. In the last design, an switched-capacitor-based DCDC boost converter with optimal efficiency tracking scheme is proposed to facilitate the self-powered health monitoring devices.Electrical and Computer Engineerin