Increasing concerns over depleting natural resources has led to the development of so-called biocomposites based on fibres from renewable resources such as flax. Although these fibres are seeing use in some applications, there is a lack of understanding concerning their processing requirements in relation to their unique physical and chemical properties. Furthermore, there is limited information regarding the links between their processing behaviour and mechanical performance. With the aim of addressing these missing links, this thesis presents a methodology for characterizing flax fibres for application in the resin infusion process and considers two important case studies with the overall goal of improving the state-of-the-art for this class of materials.Flax fibres were first characterized at the fibre level by advancing contact analysis, thermal gravimetric analysis, scanning electron microscopy and helium pycnometry. The advancing contact analysis revealed a reduction in the polar component of surface free energy after the application of silane and diluted epoxy treatments. A methodology was then developed for the characterization of the compaction and permeability of flax-based fabrics for the modelling of the resin infusion process. These parameters were quantified and used as input in a 1D process model that included capillary pressure. The model predictions for flow front evolution were shown to be in good agreement with experimental data. Alkaline treatments were shown to increase the required compaction pressure for a given porosity due to an increase in fibrillation. This had direct implications in the context of resin infusion processing due to the coupled nature of flow and compaction in this process. Consequently, a mechanical characterization revealed a decrease in flexural properties for alkaline-treated flax/epoxy composites manufactured by resin infusion due to a decrease in fibre volume fraction. A decrease in flexural properties was also noted with increasing void content.In an effort to improve the state-of-the-art for this class of materials, a case study was carried out on the incorporation of nano-modifiers in the resin infusion process. Nanocellulose was incorporated by two novel techniques; a 'grafting' method and a wet-layup method that incorporated an aqueous NC solution in the resin infusion pre-filling stage. Both methods were shown to lead to an increase in damage to the composites after subjection to a drop-weight impact event which suggested that the nano-modifier did not increase the interlaminar properties. However, an increase in interlaminar shear strength was observed by a short beam test due to an increase in fibre volume fraction as a result of softening and lubrication effects arising from the use of the aqueous NC solution.A second case study addressed the primary source of voids in a class of flax/epoxy prepregs which are generally used as a benchmark for composites manufactured by the resin infusion process. A series of compaction tests and thermal gravimetric analysis suggested that moisture and resin starvation were the primary source of voids in commercially available prepregs. Panels manufactured in an autoclave at varying pressures suggested that the latter of these issues was the dominant problem for the studied materials. The presence of voids was finally shown to lead to increased moisture sorption for flax/epoxy composites.This study stresses the coupled nature of the resin infusion process and the full implications of the use of chemical treated flax fibres. Additionally, it demonstrates the negative consequences of process-induced voids on the performance of flax/epoxy composites. It also provides useful data on the fibre surface chemistry, permeability, compaction and mechanical performance of flax-based composites. This assists in furthering the development of this class of materials with the goal of increasing their potential for use in load-bearing structures. / La préoccupation avec l'épuisement des ressources naturelles a conduit à l'élaboration des bio-composites à base de fibres renouvelables telles que le lin. Bien que ces fibres sont utilisées dans certaines applications, il y a un manque de compréhension au sujet de leurs besoins de traitement par rapport à leurs uniques propriétés physiques et chimiques. En outre, il y a peu d'information sur les liens entre leur comportement et la performance des traitements mécaniques. Dans le but de répondre à ces liens inconnus, cette thèse présente une méthodologie de caractérisation des fibres de lin pour une utilisation dans le procédé d'infusion de résine et tient compte de deux études de cas importantes dans le but d'améliorer l'état-de-l'art pour cette classe de matériaux.Les fibres de lin ont d'abord été caractérisé au niveau des fibres en faisant l'analyse de l'angle de contact, analyse thermique gravimétrique, microscopie électronique à balayage et la pycnométrie hélium. L'analyse de l'angle de contact a révélé une réduction de la composante polaire de l'énergie de surface après l'application des traitements silane et époxy dilué. Une méthode a été ensuite développée pour la caractérisation de la compression et de la perméabilité de lin à base de tissus, pour la modélisation du processus d'infusion de résine. Ces paramètres ont été quantifiés et utilisés comme input dans un modèle de processus 1D qui comprenait la pression capillaire. Le traitements alcalin a démontré une augmentation de la pression de compactage nécessaire pour une porosité donnée, en raison de l'augmentation de la fibrillation. Par conséquent, une caractérisation mécanique a révélé une baisse de propriétés de flexion pour le lin/époxy composites fabriqués par infusion de résine pour des tissus traités avec alcalines en raison d'une diminution de la fraction volumique de fibres. Une diminution des propriétés en flexion a également été notée quand le contenu de vide augmente.Dans un effort pour améliorer l'état-de l'art pour cette classe de matériaux, une étude de cas a été réalisée sur l'incorporation de nano-modificateurs dans le procédé d'infusion de résine. Du nano-cellulose a été constituée par deux nouvelles techniques; une méthode de 'greffage' et un procédé 'wet-layup' qui intègre une solution aqueuse NC avant l'infusion. Les deux méthodes ont démontré une augmentation de l'endommagement des composites après avoir été soumis à des impacts, qui suggère que le nano-modificateur n'a pas augmenté les propriétés interlaminaires. Toutefois, une augmentation de la résistance au cisaillement interlaminaire a été observée par un faisceau de test court en raison d'une augmentation de la fraction volumique des fibres à la suite d'effets de ramollissement et de lubrification provenant de l'utilisation de la solution aqueuse NC.Une deuxième étude de cas a abordé la principale source de vides dans une classe de lin/époxy préimprégnés. Une série d'essais de compactage et d'analyse thermique gravimétrique suggère que la manque d'humidité et de résine ont été la principale source de vides dans préimprégnés disponibles. Des panneaux fabriqués dans un autoclave à pression variant suggére que le dernier de ces questions était le problème dominant pour les matériaux étudiés. La présence de vides a finalement causé une dégradation des propriétés d'absorption d'humidité pour les composites lin/époxy.Cette étude souligne le caractère couplé du procédé d'infusion de résine et les implications de l'utilisation de traitement chimique des fibres de lin. En outre, il met en évidence les conséquences négatives de vides sur la performance des composites lin/époxy. Il fournit également des données utiles sur la chimie de surface des fibres, perméabilité, le compactage et la performance mécanique des composites à base de lin. Cette aide favorise le développement de cette classe de matériaux dans le but d'augmenter leur potentiel d'utilisation dans des structures portantes.

<p> In order to obtain credit for an eco-roof in building energy load calculations the steady state and time-varying thermal properties (thermal mass with evapotranspiration) must be fully understood. The following study presents results of experimentation and modeling in an effort to develop dynamic thermal mass performance metrics for eco-roof systems. The work is focused on understanding the thermal parameters (foliage &amp; soil) of an eco-roof, further validation of the EnergyPlus Green Roof Module and development of a standardized metric for assessing the time-varying thermal benefits of eco-roof systems that can be applied across building types and climate zones. </p><p> Eco-roof foliage, soil and weather parameters were continuously collected at the Green Roof Integrated PhotoVoltaic (GRIPV) project from 01/20/2011 to 08/28/2011. The parameters were used to develop an EnergyPlus eco-roof validation model. The validated eco-roof model was then used to estimate the Dynamic Benefit for Massive System (DBMS) in 4 climate-locations: Portland Oregon, Chicago Illinois, Atlanta Georgia and Houston Texas. </p><p> GRIPV30 (GRIPV soil with 30% soil organic matter) was compared to 12 previously tested eco-roof soils. GRIPV30 reduced dry soil conductivity by 50%, increased field capacity by 21% and reduced dry soil mass per unit volume by 60%. GRIPV30 soil had low conductivity at all moisture contents and high heat capacity at moderate and high moisture content. The characteristics of the GRIPV30 soil make it a good choice for moisture retention and reduction of heat flux, improved thermal mass (heat storage) when integrating an eco-roof with a building. </p><p> Eco-roof model validation was performed with constant seasonal moisture driven soil properties and resulted in acceptable measured - modeled eco-roof temperature validation. LAI has a large impact on how the Green Roof Module calculates the eco-roof energy balance with a higher impact on daytime (measured - modeled) soil temperature differential and most significant during summer. </p><p> DBMS modeling found the mild climates of Atlanta Georgia and Houston Texas with eco-roof annual DBMS of 1.03, 3% performance improvement above the standard building, based on cooling, heating and fan energy consumption. The Chicago Illinois climate with severe winter and mild spring/summer/fall has an annual DBMS of 1.01. The moderate Portland Oregon climate has a below standard DBMS of 0.97.</p>

<p> The objective of the current study is development of a coupled orthogonal curvilinear/Cartesian grid solver. The solver requires a thin orthogonal boundary layer grid and a non-uniform Cartesian grid to resolve the boundary layer on a solid surface and the flow region away from the surface, respectively. Flows inside the orthogonal boundary layer and Cartesian background grids are solved by different CFD solvers which are coupled by an overset grid method. SUGGAR code writes the grid domain connectivity information into a file that identifies grid points necessary for the overset grid interpolation. In order to satisfy mass conservation across the overlapping region, the pressure Poisson equations and the overset interpolation equations are encompassed from both of the solvers and solved simultaneously by an iterative method. </p><p> Accuracy of the coupled orthogonal curvilinear/Cartesian grid solver was evaluated in terms of flows past circular cylinders because the orthogonal boundary layer grids can be generated easily due to its simple cylindrical shape. In this study, additional numerical simulations were also performed by the original orthogonal curvilinear and Cartesian grid solvers in order to obtain the benchmark data to compare with the results of the coupled orthogonal curvilinear/Cartesian grid solver. </p><p> The coupled orthogonal curvilinear/Cartesian grid solver was applied to steady and unsteady laminar flows at Re = 40 and 200, single-phase turbulent flows at subcritical Re = 3900 and supercritical Re = 5 &times; 10⁵ and 1 &times; 10⁶, and two-phase flows at (Re, Fr) = (2.7 &times; 10⁴, 0.20), (2.7 &times; 10⁴, 0.80), and (4.58 &times; 10⁵, 1.64). Those numerical results are in good agreement with the experimental and numerical results in the literature. </p><p> Effects of the grid resolution on the numerical results were analyzed in this study. The analysis showed the more accurate resolution of near-wall regions by the boundary layer grids for the coupled orthogonal curvilinear/Cartesian grid solver. It also presented the similar trends of the flow at the subcritical Re with the vertical resolution to those observed in the literature. </p><p> The coupled orthogonal curvilinear/Cartesian grid solver predicted much delayed separations of the boundary layers at both the supercritical Re, which caused the narrower wakes and the shorter recirculation regions than those at the subcritical Re. The features of surface pressure corresponded to the postponed separations. </p><p> The solver developed in this study showed the similar trends in the two-phase flows at Fr = 0.20 and 0.80 to those observed by the past numerical studies. The trends of the vortex shedding, deviating shear layers, and the expanded wake on the free surface are more prominent in the flow at Fr = 0.80 than that at Fr = 0.20. At Re = 4.58 &times; 10⁵ and Fr = 1.64, the flow near the free surface includes the small recirculation region behind the cylinder, which corresponds to the cavity structure on the free surface in the same region, and two large symmetric recirculation regions. The shear layers separating from the cylinder surface move along the outer edges of the recirculation regions. Another pair of the shear layers is separated from the smaller recirculation region.</p>

<p> The electricity market is in need of a method to accurately predict how much peak load is removable by directly controlling residential thermostats. Utilities have been experimenting with residential demand response programs for the last decade, but inconsistent forecasting is preventing them from becoming a dependent electricity grid management tool. This dissertation documents the use of building energy models to forecast both general residential energy consumption and removable air conditioning loads. </p><p> In the models, complex buildings are represented as simple grey-box systems where the sensible energy of the entire indoor environment is balanced with the flow of energy through the envelope. When internet-connected thermostat and local weather data are inputs, twelve coefficients representing building parameters are used to non-dimensionalize the heat transfer equations governing this system. The model's performance was tested using 559 thermostats from 83 zip codes nationwide during both heating and cooling seasons. For this set, the average RMS error between the modeled and measured indoor air temperature was 0.44&deg;C and the average daily ON time prediction was 1.9% higher than the data. When combined with smart power meter data from 250 homes in Houston, TX in the summer of 2012 these models outperformed the best traditional methods by 3.4 and 28.2% predicting daily and hourly energy consumption with RMS errors of 86 and 163 MWh. The second model that was developed used only smart meter and local weather data to predict loads. It operated by correlating an effective heat transfer metric to past energy data, and even further improvement forecasting loads were observed. </p><p> During a demand response trial with Earth Networks and CenterPoint Energy in the summer of 2012, 206 internet-connected thermostats were controlled to reduce peak loads by an average of 1.13 kW. The thermostat building energy models averaged forecasting the load in the 2 hours before, during, and after these demand response tests to within 5.9%. These building energy models were also applied to generate thermostat setpoint schedules that improved the energy efficiency of homes, disaggregate loads for home efficiency scorecards and remote energy audits, and as simulation tools to test schedule changes and hardware upgrades.</p>

<p> This thesis examines the mechanical, and adhesive properties of graphene and explores using graphene as a gas separation membrane. A pressurized blister test was used to measure both the in-plane mechanical properties and adhesion energy of monolayer and few layer graphene suspended over a circular cavity in silicon oxide. The adhesion energy between graphene and silicon oxide was found to be 0.45 &plusmn; 0.02 J m^-2 for monolayer graphene and 0.31 &plusmn; 0.03 J m^-2 for samples containing two to five graphene layers. These values are larger than the adhesion energies measured in typical micromechanical structures and are comparable to solid-liquid adhesion energies. We attribute this to the extreme flexibility of graphene, which allows is to conform to the topography of even the smoothest substrates, thus making its interaction with the substrate more liquid like than solid like. In addition we found that the in-plane mechanical properties are consistent with previously reported values.</p><p> We also show that ultraviolet-induced oxidative etching can create pores in micrometer-sized graphene membranes, and the resulting membranes can be used as molecular sieves. A pressurized blister test, similar to that used for testing the mechanical properties, and mechanical resonance are used to measure the transport of a range of gases (H₂, CO₂, Ar, N₂, CH₄, and SF₆) through the pores. The experimentally measured leak rate, separation factors, and Raman spectrum agree well with models based on effusion through a small number of angstrom-sized pores.</p><p> Lastly, we work toward creating large scale gas separation membranes from chemical vapor deposition (CVD) grown graphene films. CVD graphene films are grown on copper foils and transferred to a polymer support or suspended over openings in copper. Films are measured in a time lag permeation apparatus to get gas permeation and ideal gas separation factors.</p>

<p> This work investigates bi-phase flows in a series of experimental and computational studies with a tribological focus. Three silver-based complexes are chemically designed, through collaboration with the Marks group, for use as additives in high-temperature lubrication. Mixtures of engine oil and various concentrations of these nanoparticle complexes lubricate sliding steel surfaces in ball-on-disk tests. Friction and wear measurements demonstrate that the silver complexes provide beneficial wear reduction over a range of temperatures and loads. The silver pyrazole-pyridine complex in particular provides excellent friction performance at high temperatures (> 200&deg;C) and high concentration (20 wt%). In a similar set of experiments, hexyltrimethoxy silane is added to polyalphaolefin (PAO) oil contaminated with sand. As debris, dirt, and dust are commonly present in lubricants, it is desirable to find additives which lessen the abrasive damage of these hard contaminant particles. Chemical reaction between the sand and organosilane is found to be low, which results in minimal improvements in friction and wear. A tin catalyst is recommended to improve the reactivity of the sand and organosilane for future testing. Finally, a computational model of fluid-solid flow is developed to examine the characteristics of particle-laden viscous flows. Existing computational techniques utilizing a distributed Lagrange multiplier (DLM) method and a mechanistic collision model are extended for multi-particle collisions and elastohydrodynamic lubrication, accounting for surface deformations through the solution of elasticity equations. This deterministic numerical model allows for various sizes and shapes of particles, as well as dilute to dense suspensions in viscous, laminar flow. Furthermore, the model is computationally fast and applicable to a wide range of thin-film flows. The numerical model is utilized in a study of particle-fluid flow through narrow channels with wall features of increasing size. Particles are found to have different equilibrium positions based on the flow Reynolds number, with minimal effects from the wall features. In addition, results for rigid particle motion entrapped in a deformable channel are presented as a preliminary investigation. The location of maximum surface pressure and deformation is predicted to shift slightly from the point of contact in the direction of motion of the moving particle.</p>

A study is made of the motivation involved in the use of Fourier integrals in the formulation of solutions to certain radially symmetric elasticity problems. A new class of such problems is studied: that of stresses in a half space penetrated by a circular cylindrical hole with axis normal to the bounding plane. It is demonstrated by solution of an example that this class of problems may be solved using a large, high-speed computing machine. A problem related to stresses around an oil well is also studied.

The Prandtl-Meyer flow of a pure, monatomic gas, including the effects of ionization and recombination, has been numerically calculated. Expressions describing the thermodynamic state of the gas mixture were obtained from statistical mechanics, the components making up the mixture being considered as perfect gases. Expressions describing the kinetic behavior of the ionizing-recombining gas mixture, using J. J. Thomson's three-body recombination coefficient, and expressions describing the gas dynamic behavior of the reacting gas mixture were derived. An iterative method of solving the equations governing equilibrium flow was derived. Solution was carried out, using an IBM 1620 Data Processing System, for a particular initial state: degree of ionization = .9000, dimensionless temperature = .1250 (=35,650&deg;K in Helium), Mach number = 1.000. Curves showing the variation of the state and motion variables obtained are presented. Using the assumptions that the pressure distribution along a streamline and the shape of a streamline in a non-equilibrium Prandtl-Meyer flow are the same as for an equilibrium flow with the same initial conditions, the governing flow equations were reduced to two, coupled, ordinary differential equations in the degree-of-ionization and velocity distributions along a streamline. Solution of these equations, using the pressure distribution and streamline shape obtained from the corresponding equilibrium flow, gives the degree of ionization and velocity along a particular streamline as functions of angular position in the expansion wave. Numerical solution of these equations, using an IBM 1620 Data Processing System, was carried out for flow from an initial state the same as the above equilibrium solution, along six streamlines ranging from very near to the corner (.086 cm in Helium) to very far from the corner (86 cm in Helium). Curves showing the variations of the state and motion variables obtained are presented.

A finite element method is presented for the three (3) dimensional, dynamic analysis of the offshore pipelaying problem. A geometrically nonlinear beam element and elastic bi-linear support elements are utilized to simultaneously model the pipeline, stinger, pipe supports and seabed. Numerical integration is used to obtain a time domain solution for the dynamic response of the pipeline. An iterative Newton's method procedure is used to solve the nonlinear system of algebraic equations, produced by the method, at each timestep. Both the end force equations and stiffness, mass and damping matrices for the beam element are given. The method presented offers several advantages over existing techniques. There are no theoretical limitations on the magnitude of the pipe displacements. It is not assumed that the dynamic response of the pipeline is linear or that it represents a small perturbation about the static solution. The beam element developed is completely general and can be used to model redundant structures. It provides for both the torsional deformation and elongation of the pipeline, and permits the use of different physical properties in each principal direction. Although the element is linearly elastic, it can be extended to nonlinear elasticity by the inclusion of a suitable moment curvature relationship. The results of numerical calculations are presented for a typical offshore pipelaying problem.

Nondestructive evaluation (NDE) is the inspection of samples for corrosion and physical defects without altering them in any way. NDE has a critical role in the robotic inspection of energy pipelines in order to prevent catastrophic failures. Magnetic flux leakage (MFL) sensing is by far the most effective technique for robotic inspection of ferromagnetic pipes and tubular specimens. Defect detection using MFL sensing is a mature area of work, but defect characterization using MFL sensing is an open research problem. Several issues involved in this process are not well understood---for example, the interplay of the components of the 3-dimensional MFL field for 3-dimensional defects, the spatial properties of the MFL field components, the effect of sensor lift-off on MFL signals, and the relationships between defect properties and MFL signal properties. This dissertation addresses these issues using a systematic approach. First the MFL sensing problem is decomposed into the forward and inverse problems. Subsequently, a tractable forward model is presented which is capable of predicting the 3-dimensional MFL field of a known 3-dimensional surface-breaking defect. Important properties of the MFL field and their correlation with sensing parameters and defect parameters are established using the model and simulation. A linear inversion technique is presented which exploits the structure and properties of the forward model to characterize defects based on measured MFL signals. This dissertation also proposes a general framework to solve the inverse problem independent of the NDE modality in use. This framework uses the principles of data fusion and neural networks, and is illustrated using both MFL signals as well as another NDE technique, namely ultrasonic testing. Finally, this dissertation addresses the problem of pipe wall thickness measurement as a special case of the inverse problem and develops a novel technique for pipe wall thickness measurement using MFL signals.