DSpace@RPI (Rensselaer Polytechnic Institute)
Not a member yet
6809 research outputs found
Sort by
Medium, matter, form, and process: dynamic hygrothermal polymeric membranes
No full textAugust 2014School of ArchitectureExisting knowledge of building envelope system performance and technology for processing humidity is embedded within, and biased by, a historic lineage of attempting to hermetically seal and mechanically control the passage of humidity into building interiors. This paradigm perpetuates an energy consumptive process, limits the ability for distributed user control of the thermal environment, and is also problematic to the extent that it may cause increased symptoms of sick building syndrome with an unbalanced psychrometric profile, poor ventilation, lack of natural daylight, and microbial contamination, through the reduction of protective microbial biodiversity. In addition, previous studies of mechanical desiccant systems fail to consider an integrated dehumidification function distributed within the building envelope system itself and rather emphasize a trajectory of improving isolated system efficiencies without consideration of holistic building design impact. Without consideration of a cultural shift with the potential of emerging material technologies for building envelope design, existing building technologies for hot-humid and hybrid moderate climates lack the effectiveness to simultaneously address environmental phenomena of humidity, light, heat, ventilation, and water recu-peration. If hydrogels are integrated into the design of exterior and interior building envelopes, then an effective dehumidification function could simultaneously be achieved in a multifunctional systemic approach that includes water recuperation, dynamic thermal capacitance, light diffusion, and ventilation actuation properties. This thesis proposes the design of hygrothermal polymeric membranes towards effective multi-scalar responsiveness to dynamic environmental conditions around and through building envelopes, including relative humidity, dry-bulb temperature, solar radiation, natural ventilation, and water recuperation regeneration cycles. This analysis is achieved by acquiring performance data from a series of experiments determining material behaviors, and crosslinking that data with a limited series of digital building-scale energy models in order to determine the viability of replacing existing models of humidity management with the proposed model of distributing intelligent hydrogel-based desiccant materials within building components such as shading assemblies and radiant structural components. These analytical models provide a foundational framework that can inform future research directions for building-integrated desiccant materials, particularly for hot-humid climates, especially with regards to energy conservation, adequate ventilation, microbial management, and water recuperation. The dynamic hygrothermal polymeric membrane research initiates a cultural shift in building envelope design by considering the dehumidification function as integral to the building envelope system and by conceiving of both exterior and interior environmental phenomena through the membrane as an interdependent ecosystem, effectively resulting in improved energy performance, improved natural daylighting, improved water conservation, improved ventilation quality, and improved thermal comfort through distributed control. These initial baseline studies inform prospective prioritization of environmental performance criteria for dynamic hygrothermal polymeric membranes with consideration of the influence of material limitations, economic limitations, and social implications. The design-based research is lensed through a theoretical framework of four causes in the design process: medium, matter, form, and process. Theoretically, the design research emerges from an analysis of the phenomena of medium (humidity), within an ecological imperative. This foundation for ecologically-based design research is balanced by an intention of reflexive sustainability, which encompasses aspects of subjectivity and cultural implications.Ph
Arcadian pasts and futures: making and breaking a lignite coal landscape in southern greece
No full textAugust 2023School of Humanities, Arts, and Social SciencesAt the end of the 1950s, Greece’s state-owned electrical utility the Public Power Corporation (PPC) surveyed large reserves of a low-quality energy resource called lignite coal around the town of Megalopolis. Since the inauguration of the lignite mines and power plant in 1970, lignite coal exploitation has organized the social, economic, and technical shape of Megalopolis—a lignite landscape. Since the announcement of a nation-wide “delignification” program in 2019, Megalopolis’s lignite landscape has unwound as government neglect and private wind energy developers carelessly intercede near important natural and cultural heritage sites. However, by dismantling the lignite landscape, the energy transition at Megalopolis is not only revealing of an increasingly uncertain future but has unmoored the area from the historical and temporal relations enacted by the large-scale landscape practice of lignite mining, transportation, and burning. Drawing on studies on resources, energy transition, landscape, as well as historicity and temporality, I investigate the meaning of Greece’s energy transition and describes the consequent ways people in Megalopolis are imagining historical meaning and feeling temporal experience. Methodologically, I combine participant observation on the Mount Lykaion Archaeological excavation, ethnographic interviews with local residents from the Megalopolis basin and Lykaion mountains above, archeological labors and lignite miners, Arcadian archeological and heritage experts, and local mountaineers and environmentalists, with archival research at the historical archives of Greece’s Public Power Corporation and Arcadian’s archeological ministry. Overall, this research shows how fossil fuel and energy systems and the renewable energy transition shape, and are shaped, by the values and meanings of the past, present, and future.Ph
Are Human Explanations Always Helpful? Towards Objective Evaluation of Human Natural Language Explanations
No full textHuman-annotated labels and explanations are critical for training explainable NLP models. However, unlike human-annotated labels whose quality is easier to calibrate (e.g., with a majority vote), human-crafted free-form explanations can be quite subjective, as some recent works have discussed. Before blindly using them as ground truth to train ML models, a vital question needs to be asked: How do we evaluate a human-annotated explanation's quality? In this paper, we build on the view that the quality of a human-annotated explanation can be measured based on its helpfulness (or impairment) to the ML models' performance for the desired NLP tasks for which the annotations were collected. In comparison to the commonly used Simulatability score, we define a new metric that can take into consideration the helpfulness of an explanation for model performance at both fine-tuning and inference. With the help of a unified dataset format, we evaluated the proposed metric on five datasets (e.g., e-SNLI) against two model architectures (T5 and BART), and the results show that our proposed metric can objectively evaluate the quality of human-annotated explanations, while Simulatability falls short
A framework for modeling complex integrated building systems at whole-building scale with co-simulation: applied to a coupled simulation between a facade system model and a whole building energy model
No full textMay 2018School of ArchitectureWithin the context of the designing and constructing systems for the built environment, the increasing complexity of environmental and ecological systems requirements has led to the widespread adoption of building energy modeling (BEM) across the architecture and engineering (AE) industry. Within this inquiry, the requirements of advanced integrated facades (AIFs) are examined as use cases for the development of novel computation techniques to model and quantify building integrated energy capture technologies that could support significant progress towards clean on-site energy self-sufficiency. An AIF may add a multitude of benefits to the building such as clean on-site energy, moderated heat gain, daylighting, increased visual comfort, and improved human comfort. Reliable methods to simulate the impact of these complex technologies at whole-building scale are not currently available to building designers. Using conventional BEM to model the impacts of the integrated systems on whole-building energy metrics is either too time consuming with a project’s timeline or infeasible with the modeling tool’s prescribed functionality. Typical BEM methods, in order to facilitate access and ease of use, are developed for whole-building simulation at an annual time-scale, therefore they are too coarse-grained to accurately model the simultaneous interactions of an AIF, especially when combined with ambient energy capture technologies. A method called co-simulation is emerging across multiple fields as a standard for coupling models developed within different modeling environments. In buildings research and engineering, computational system models are developed to quantify the transport phenomena of a building system that would otherwise be infeasible to simulate with BEM or CFD. Co-simulation provides a method for building energy analysts to more readily study emerging integrated building systems using computational system models coupled with conventional BEM tools, like EnergyPlus. Leveraging co-simulation features developed into EnergyPlus, this research proposed the use of the method to model and quantifying two emerging AIFs using whole-building energy metrics. Within the scope of this thesis, a novel model has been co-developed, called the Modular Network Model, that is capable of modeling AIFs by discretizing the building envelope systems into repeatable modules, which are combined using a network model method and balanced using conservation of mass and energy to solve for the simultaneous transport phenomena and the systems’ interaction with the building. As a proof-of-concept test, the Modular Network Model was applied to model the Building envelope-Integrated, Transparent, Concentrating Photovoltaic and Thermal (BITCoPT) system at whole-building scale, capable of modeling the simultaneous energy and mass transport phenomena necessary to quantify the photovoltaic electrical generation, thermal energy collection, modulated solar heat gain, and cavity thermodynamics. The Modular Network Model of the BITCoPT system was coupled and co-simulated with EnergyPlus, an industry standard BEM software, thereby allowing designers to more readily study the impact of the system on whole-building energy metrics, while reducing model development time. The process was repeated for the EcoCeramic Envelope System (EES), whereby a computational system model was developed within the Modelica system modeling language, following the same modeling structure of the Modular Network Model. The EES system model was developed to model the interactions between the outdoor air temperature, wind, and radiation with the ceramic system and thermal fluid cavity. The EES model was exported as a Functional Mock-up Unit (FMU) and co-simulated within EnergyPlus as a thermally adaptive building envelope with solar hot water collection that provided energy to building heating. Co-simulation between the AIF system models and building energy models improved the predictions of the systems’ impact on building energy consumption metrics and heat transfer performance as demonstrated through the comparative analysis shown in Chapters 3, 4, and 5.Ph
Optimal freight pricing considering routing and the ordering behavior of receivers
No full textMay 2023School of EngineeringUrban deliveries are crucial for the development of cities; however, they produce negative externalities, such as emissions and congestion. At first glance, it may appear carriers are generating these externalities, but in reality, the ordering decisions of the receivers of supplies, i.e., commercial establishments, play a significant role in creating freight externalities. Receivers create the need for freight trips. Therefore, policies that foster changes in the receivers' ordering decisions, such as reducing the number of orders, are necessary to reduce externalities produced by freight activity for the benefit of society.These policies should not only consider the welfare of society but also balance the profit-maximizing nature of businesses. For example, businesses are often limited by the available floorspace and must decide how to allocate this space between productive activities and storage. The allocation of floorspace affects their ordering decisions and, ultimately, their profits. This research contributes to promoting sustainable urban deliveries by developing policy procedures that induce changes in receivers' behavior. To this effect, this research is the first to develop an analytical formulation for the Social Optimal Routing and Ordering Problem (SOROP) that considers: (i) the profit-maximizing behavior of receivers, (ii) an externality charge to account for the externalities produced by their ordering decisions, and (iii) the effect of these decisions in the routing structure. The model clusters receivers based on their ordering patterns, grouping those who order with similar frequency into the same delivery tour.
The results of a numerical experiment reveal that imposing an externality charge on receivers reduces the number of orders placed, increases order sizes, and reduces negative externalities generated by urban deliveries. Furthermore, the study includes an elasticity analysis to identify the effect of cost components and establishment attributes on the receivers' ordering behavior. This research demonstrates that receiver pricing plays a crucial role in urban freight management as it forces the receivers to internalize the externalities produced, ultimately reducing the number of freight trips.Ph
Polarized transport requires ap-1-mediated recruitment of kif13a and kif13b at the trans-golgi
No full textAugust 2023School of ScienceThe nervous system is comprised of neurons, which send and receive electrochemicalsignals. To facilitate this function, they possess an exotic morphology containing
complementary extensions: the signal-sending axon and signal-receiving dendrites.
Unique complements of membrane proteins facilitate these functions, making neurons
polarized. Membrane trafficking maintains this polarity but is poorly understood.
Dendritically polarized membrane trafficking requires the sorting of dendritic proteins
into vesicles at the Golgi, which must recruit the correct molecular motors to confer
transport to the dendrites. The related kinesin-3 molecular motors, KIF13A and KIF13B,
transport dendrite-selective vesicles. Whether there are differences between these
motors and their binding and transport of dendrite-selective vesicles is unknown. I used
quantitative fluorescence imaging to determine that KIF13s differ in their colocalization
and cotransport with dendrite-selective transferrin receptor (TfR) vesicles: KIF13A is
specialized for dendrite-selective transport, with KIF13B assisting but additionally
transporting axon-selective vesicles containing neuron-glia cell adhesion molecule
(NgCAM). I found that both motors are recruited to Golgi-derived vesicles at the transGolgi network (TGN) by binding the heterotetrameric clathrin adaptor protein (AP)
complex-1. Critically, disrupting this interaction reduces dendrite- and axon-selective
transport of TfR and NgCAM. However, AP-1 does not serve as the long-term kinesin
adaptor for either vesicle population. In this model, KIF13s mediate polarized transport of
distinct and overlapping vesicle populations, with AP-1 performing initial recruitment of
KIF13s to these vesicles at the TGN.
This project was only possible because of the development of a novel method to
visualize organelle-bound kinesins. Before this method, researchers usually expressed
full-length kinesins fused to a fluorophore that provided poor and inconsistent organelle
labeling. Often there would be a soluble population of kinesin that likely concealed
kinesin-bound organelles. The current model argues that this soluble pool consists of
inhibited kinesins whose motor domains are bound to their cargo-binding tails.
To overcome this problem, we expressed only the cargo-binding tails of the kinesins
implicated in neuronal vesicle transport. For the smaller Kinesin-1s, we controlled
transcription by incorporating a nuclear localization signal and zinc finger domain in the
tail constructs. Kinesin-1 motors not bound to organelles would enter the nucleus where
they would bind the plasmid and halt transcription. This strategy, combined with short
expression times, gave a prodigious improvement in vesicle labeling for many transport
kinesins. We now possess a powerful method to directly observe the organelles that
kinesins bind.Ph
Interfacial phenomena in heat pipes: microscopic and molecular dynamics study
No full textMay 2023School of EngineeringMicro-scale transport phenomena have gained prominence in various high heat flux dissipation devices in microelectronics, where the hotspots created require that each chip be cooled individually. Heat pipes have been an attractive choice for high heat flux operations because of their compact, low-maintenance design. Heat pipes are passive heat transfer devices that transport energy from one end to the other based on capillary and Marangoni flows. Temperature variation along the length of the heat pipe together with a gradient in surface tension, cause the fluid to recirculate between the evaporator and condenser ends. Heat pipes have found application in space research, where extremely small devices are needed for an efficient cooling system. Understanding phase-change heat and mass transfer in the contact line region is important for optimizing many industrial and biological processes like nucleate boiling, spreading, coating, self-assembly, evaporation, condensation, and tear films. The three-phase contact line is the micro-region where solid, liquid and vapor phases coexist. The fluid flow and heat transfer occurring in a thin-film depend on the shape of the contact line. Thus, understanding the intermolecular force field, which eventually governs the meniscus shape and curvature is important. Our aim is to study phase-change phenomena in this 3-phase contact line by obtainingthe interfacial characteristics at the meniscus of a thin evaporating film. A finite element model was previously developed that simulates the fluid dynamics and heat transfer in a thin liquid meniscus on a solid substrate by solving a pair of partial differential equations. After reaching a steady-state, the outside solid wall temperature was oscillated with different
frequencies and varying amplitude. The effects of oscillation on the liquid film thickness and heat transfer at the solid-liquid boundary were studied. It was observed that the liquid film also oscillates and follows the wall temperature for lower frequencies of oscillation. However, for higher frequencies, oscillations in the film thickness were damped out. The heat flux profiles at the solid-liquid boundary show that there is no optimal oscillation frequency at which the evaporative heat flux is maximum. Knowing what the optimal frequency is will offer better control in designing of a heat transfer device/process, giving the maximum heat transfer efficiency. A heat-transfer cell which houses a 4-inch wafer has also been designed to study the 3-phase contact line experimentally. To validate the theoretical predictions, Hamaker constant was calculated by performing isothermal experiments on silicon surface. By modifying the surface in a simple way and studying the thin-film evaporation on it, we hope to achieve a better understanding of the effect that solid-liquid interactions have on the heat transfer. We discuss the effect that fluid properties have on the behavior of the evaporating liquid and the oscillation of the film. The overall heat flux and amplitude of oscillation of thin film thickness over one cycle were compared across different fluids. The latent heat of vaporization of the liquid and the Hamaker constant of the system were found
to control the oscillating thickness of the liquid film on solid. The total heat transfer also depends on the thermal diffusivity of the liquid and solid system. The study could further be extended for a fluid partially wetting a solid surface by modifying the disjoining pressure isotherm. The short-range (polar) contribution to the disjoining pressure alter the interaction between the liquid and solid molecules. Pentane is a commonly used fluid in heat pipes in microgravity. It was reported in our group that using a binary mixture of alkanes (94% pentane and 6% isohexane) in a constrained vapor bubble (CVB) experiment on International Space Station (ISS) improved the performance of the heat pipe as compared to using pure pentane as the working fluid. The change in mixture composition and temperature along the length of the heat pipe caused opposing effects on the surface tension, essentially reducing the Marangoni stress. We studied the liquid-vapor equilibrium of a mixture of n-pentane and 2-methylpentane (isohexane) at a molecular scale using the software GROMACS for running molecular dynamics simulations. Vapor-liquid equilibrium simulations were carried out for two single-component systems (n-pentane and 2-methylpentane) and 3 binary mixture compositions (molar ratio 25:75, 50:50, and 75:25) of pentane and 2-methylpentane. Our objective is to study these mixtures at a molecular level to provide insight into the intermolecular interactions between these alkanes which are important for knowing the properties of the evaporating meniscus. We discuss the structure, thermodynamics and dynamics of the binary mixtures. The surface tension was found to vary linearly with the mixture composition indicating a near-ideal behavior of the binary mixture. The surface tension also decreases with an increase in temperature. Studying the orientations of the molecules revealed that the liquid-vapor interface could be divided into three regions corresponding to different orientations, however, the preference for any orientation is only slight. The isohexane molecule was observed to orient itself such that its bulky end points to the liquid bulk. Molecular dynamics simulations will help in understanding the structure and dynamics of the mixture and aid in better designing the operating parameters of heat transfer devices to improve their effectiveness.Ph
Designing learning control algorithms for mechanical systems with complex dynamics
No full textAugust 2023School of EngineeringNumerous mechanical processes are governed by complex multi-scale dynamics (e.g. manufacturing, robotic systems, and building thermal control), which often involve the interaction of various physical phenomena, making it difficult to derive suitable models for controller design purposes. In all of these systems however, the opportunity is that there is a massive increase in available data, with high-fidelity simulations, improved experimental measurement techniques, and increased computational power. Learning control techniques can leverage this data, offering the potential to overcome the challenges posed by complex dynamics in mechanical systems and optimize control strategies without relying solely on accuratephysics-based models. The work in this dissertation is a demonstration of the endeavors to develop intelligent control strategies for mechanical systems, particularly with complex dynamics, through the investigation of learning-based control strategies in two representative applications, namely (1) passive building thermal control and (2) metal-additive manufacturing (laser powderbed fusion). These applications have been chosen due to their inherent complexity, as the dynamic behavior of these systems is spatially and temporally distributed, governed by principles of mass and energy conservation, as well as phase changes, posing challenges in effective controller design. However, they also present a significant opportunity to improve
control performance and system efficiency through the implementation of learning control approaches. For the first application, passive building thermal control, this thesis presents a robust control strategy that harnesses climatic resources to minimize the use of mechanical heating/cooling energy. The algorithm is developed with a focus on reducing training efforts and enhancing the adaptability of the learning control algorithm. To achieve this, first, an approach to incorporate domain knowledge in the form of an expert is introduced, followed by the design of the learning control algorithm. The expert is then used to assist the initialization of the learning control algorithm, such that the controller mimics the behavior of the expert to serve as baseline, accelerating the training process and enhancing the initial performance. The expert-assisted initialization also serves as a means to reduce undesirable behavior and enhance the performance in the final controller. Next, the baseline controllers are deployed in contrasting climates to further learn a representative control strategy for each climate. The developed learning control algorithms are tested in similar yet unforeseen climate and building conditions, in which we find that the control algorithms are able to adapt to climate and building-specific variations. The expert-assistance also proves to be an efficient means to reduce training efforts, and greatly enhance the performance of the developed controllers. In the second application, metal-additive manufacturing, a control strategy to effectively regulate the measurements from the process is presented. Methods to safely develop a learning control algorithm that is capable of robust control in the physical system are investigated. On pursuit of this goal, a sim-to-real (simulation-to-reality) learning control approach is proposed, involving three key steps. First, a physics-informed model is developed to replicate the system dynamics and serve as a basis for training the learning control algorithm. This model provides insights into anticipated measurement deviations and enables the controller to respond to prior deviations that can be compensated. Subsequently, the algorithm is trained using the developed model, reducing training efforts and ensuring the safe development of the control strategy. Finally, the trained control strategy is deployed in the physical system, and its performance is evaluated under various build conditions. Upon experimental deployment, we find that the algorithm is applicable to novel build geometries without further tuning or modification, showcasing the geometry-informed capabilities of the algorithm. Furthermore, the sim-to-real approach serves as an competent strategy for the mitigation of training time and safety issues.Ph
The molecular characterization of tau glycoprotein/glycan interactions and phosphorylation in alzheimer’s disease
No full textJune 2023School of ScienceAlzheimer’s disease (AD) is the 7th leading cause of death in the United States as of 2021, with poorly understood etiology and limited therapeutic measures. Of the various hypotheses for AD pathogenesis, herpes virus has grown as an attractive contributing agent due to a myriad of correlative evidence. Herpes virus shares an intriguing connection to one of the pathological hallmarks of AD, hyperphosphorylated Tau; both are internalized by neurons via a 3-O-sulfated heparan sulfate (3-O-S HS) chain on heparan sulfate proteoglycans. These are two of the only eight known proteins to specifically recognize the 3-O-S moiety on HS. However, to investigate the potential ternary interaction of AD-related Tau, herpes virus, and heparan sulfate, we must first define each binary interactions of our biomolecular triad, as two of these interfaces, herpes viral glycoprotein D and heparan sulfate, and phosphorylated Tau and heparan sulfate, are not fully understood. Herein, we present here the results of investigations into gD-HS interface, the ternary interaction between gD-HS-Tau, and the impact of phosphorylation on the Tau-HS interface. For the gD-HS interface, we utilized various biophysical assays to clarify the protein regions involved in HS binding and to investigate the glycan preferences of gD. We found that the first 22 amino acids of gD are disposable for HS binding, contrary to previous literature. We also show that gD has a strong preference for 6-O-sulfation, a variable requirement for 2-O-sulfation, and recognizes 3-O-sulfation using SPR competition assays, SV-AUC, and glycan microarray. Furthermore, a de-glycosylated mutant of gD revealed a different dissociation mode when interacting with heparin, pointing to the involvement of glycosylation in the gD-HS interaction. SPR, affinity chromatography, and SV-AUC provide indirect and direct evidence of a ternary complex between full-length Tau, gD285, and an HS analog heparin. Phosphorylation is the most important post-translational modification of Tau in Alzheimer’s disease. We accomplished phosphorylation of Tau by co-expressing a kinase, glycogen synthase kinase 3β (GSK3β) and Tau in E. coli. Phosphorylation by GSK3β did not impact Tau’s affinity for heparin, a sulfated analog of heparan sulfate, as shown by SPR kinetic assays and heparin affinity chromatography. However, phosphorylation did change the conformational dynamics of Tau and its complex with heparin as shown by SV-AUC. Furthermore, SPR competition assays and glycan microarray revealed that phosphorylation shifted the glycan preferences of Tau away from 3-O-S. This may have implications for the formation of ternary complexes of HS, pTau, and other proteins in AD pathogenesis. This dissertation provides detailed characterization on two important protein-glycan interactions, glycoprotein D/heparan sulfate and phosphorylated Tau/heparan sulfate, and investigates a ternary complex between glycoprotein D/heparan sulfate/Tau which has implications for Alzheimer’s disease, herpes viral entry, and drug development.Ph
Eat4Genes: a bioinformatic rational gene targeting app and prototype model for improving human health
No full textIntroduction and aimsDietary Rational Gene Targeting (DRGT) is a therapeutic dietary strategy that uses healthy dietary agents to modulate the expression of disease-causing genes back toward the normal. Here we use the DRGT approach to (1) identify human studies assessing gene expression after ingestion of healthy dietary agents with an emphasis on whole foods, and (2) use this data to construct an online dietary guide app prototype toward eventually aiding patients, healthcare providers, community and researchers in treating and preventing numerous health conditions. MethodsWe used the keywords "human", "gene expression" and separately, 51 different dietary agents with reported health benefits to search GEO, PubMed, Google Scholar, Clinical trials, Cochrane library, and EMBL-EBI databases for related studies. Studies meeting qualifying criteria were assessed for gene modulations. The R-Shiny platform was utilized to construct an interactive app called "Eat4Genes". ResultsFifty-one human ingestion studies (37 whole food related) and 96 key risk genes were identified. Human gene expression studies were found for 18 of 41 searched whole foods or extracts. App construction included the option to select either specific conditions/diseases or genes followed by food guide suggestions, key target genes, data sources and links, dietary suggestion rankings, bar chart or bubble chart visualization, optional full report, and nutrient categories. We also present user scenarios from physician and researcher perspectives. ConclusionIn conclusion, an interactive dietary guide app prototype has been constructed as a first step towards eventually translating our DRGT strategy into an innovative, low-cost, healthy, and readily translatable public resource to improve health