In order to improve the state of turbulence modeling for three-dimensional flows, more detailed information on the fundamental physics of the flow is required. It has been recognized for some time now that organized motions or coherent structures in the flow play a large part in determining the flow characteristics, and there is now a large body of literature dealing with various aspects of coherent structures. However, almost all of the existing literature deal with mean two-dimensional flows with very little reported for mean three-dimensional flows. In the present study, measurements were performed in a three-dimensional, pressure-driven turbulent boundary layer (<i>Re</i>_θ = 5936) in the flow around a wing-body junction with a variety of multiple-sensor probes, to examine the features of the coherent structures in the flow. This test flow has a number of practical applications and was selected because of its strong three-dimensional nature and the availability of an extensive set of mean-flow measurements from previous investigations. The measurements were carried out with a hot-wire rake with sixteen sensors spaced approximately logarithmically over 25.4 mm (1 inch), a parallel-sensor probe with two parallel sensors spaced approximately 4.8 mm apart, a rotatable wall-sensor probe with two wall-mounted hot-film sensors spaced 6.93 mm apart and a traversable wall-sensor probe with two variable-spacing wall-mounted hot-film sensors. The hot-wire rake was used to examine the structure of the flow in both the Y (normal to the wall) and Z (spanwise) directions. The parallel and rotatable wall-sensor probes were used to look at the angular characteristics of the coherent structures in the flow and at the wall, respectively, and the spanwise structure of the flow at the wall was examined through the traversable wall-sensor probe. The results of the measurements show that the spectral characteristics of the flow are affected by three-dimensional effects. The direction of motion of the coherent structures lags behind the local mean-velocity vectors in the X-Z plane (parallel to the wall) with very little variation with frequency (structure size). Unlike two-dimensional boundary layers, the spectral variation of the convective wave speed does not collapse when normalized with the local mean velocity and friction velocity in the outer and inner regions, respectively. In the outer region of the boundary layer, the distribution of the intermittency with Y appears to agree quite closely with previously reported results for two-dimensional boundary layers. The mean ejection frequency in the near-wall flow and the frequency at the peak of the first moment of the wall shear-stress power spectrum show fairly close agreement, consistent with previously reported results for a two dimensional boundary layer. The measurements with the traversable wall-sensor probe indicate the presence of an organized structure, probably low-speed streaks in the near-wall region, with a preferred spanwise spacing. This spanwise spacing was found to be Î Î * = 85 and 135 at two different measurement stations. somewhat different from the well accepted value of Î Î * = 100 for two-dimensional boundary layers. Time-delayed correlations of the velocity signal over a range of Y locations reveal an inclined linear wavefront similar to previously reported results for a two-dimensional boundary layer. / Ph. D.

<p> An implicit finite difference technique has been developed for the solution of the steady two dimensional boundary layer equations. </p> <p> The numerical method is free of stability limitations and similarity assumptions. Use has been made of Wu-type starting profiles which enable one to start the calculation from the leading edge. </p> <p> Attractive features of the technique are its simplicity, flexibility and applicability to a wide range of boundary layer problems. In addition, results obtained from several case studies indicate that the numerical procedure is both accurate and fast. </p> / Thesis / Master of Engineering (MEngr)

Some important Computational Fluid Dynamics (CFD) applications and flow solvers are strongly influenced by the effects of low Mach number and/or high heat transfer rates. The present study focuses on the validation of a new flow solver over a range of Mach numbers and heat transfer rates of practical interest. The validation study addresses both algorithm performance via convergence rate and accuracy via comparisons with an analytical similarity solution, for flow past a flat plate held at a constant temperature. The ranges of flow conditions investigated in this study are (0.1 <= Mref <= 3.0) and (0.1 <= Twall <= 10.0). The algorithm has been found to converge well for most flow conditions tested. Optimal convergence rate is more strongly influenced by choice of time step in subsonic flows than in supersonic flows. Validation comparisons show the algorithm to maintain a consistent and acceptable level of accuracy for most tested flow speeds and wall temperatures, with most deviations attributable to flow physics constraints imposed by the theoretical result or to grid resolution. One possible exception, where the outer range of capability of the present algorithm may have been reached, is noted.

No description available.

No description available.

No description available.

No description available.

Natural composites like wood are scale-integrated structures that range from molecular to the macroscopic scale. Inspired by this design, layer-by-layer (LbL) deposition technique was used to create lignocellulosic composites from isolated wood polymers namely cellulose and lignin, with a lamellar architecture. In the first phase of the study, adsorption of alkali lignin onto cationic surfaces was investigated using a quartz crystal microbalance with dissipation monitoring (QCM-D). Complete coverage of the cationic surface with alkali lignin occured at low solution concentration; large affinity coefficients were calculated for this system at differing pH levels. Adsorption studies with organosolv lignin in an organic solvent, and spectroscopic analysis of mixtures of cationic polymer with alkali lignin revealed a non-covalent interaction. The work demonstrated how noncovalent interactions could be exploited to molecular organize thin polyphenolic biopolymers on cationic surfaces. The second phase of the study examined the adsorption steps during the LbL assembly process to create novel lignocellulosic composites. LbL assembly was carried out using oxidized nanocellulose (NC) and lignin, along with a cationic polymer poly(diallyldimethylammonium chloride) (PDDA). QCM-D was used to follow the sequential adsorption process of the three different polymers. Two viscoelastic models, namely Johannsmann and Voigt, were respectively used to calculate the areal mass and thickness of the adsorbed layers. Atomic force microscopy studies showed a complete coverage of the surface with lignin in all the disposition cycles, however, surface coverage with NC was seen to increase with the number of layers. Free-standing composite films were obtained when the LbL process was carried out for 250 deposition cycles (500 bilayers) on a cellulose acetate substrate, following the dissolution of the substrate in acetone. Scanning electron microscopy of the cryo-fractured cross-sections showed a lamellar structure, and the thickness per adsorption cycle was estimated to be 17 nm. The third phase of the study investigated the effect of LbL ordering of the polymers versus a cast film composed of a blended mixture of the polymers, using dynamic mechanical analysis. A tan ï ¤ peak was observed in the 30 – 40 ºC region for both films, which was observed in the neat NC film. Heating of the samples under a compressive force produced opposite effects in the films, as the LbL films exhibited swelling, whereas the cast films showed densification. The apparent activation energy of this transition (65 – 80 kJ mol-1) in cast films, calculated based on the Arrhenius equation was found to be coincident to those reported for the ï ¢ transition of amorphous cellulose. The peak was seen to disappear in case of LbL films in the second heat, whereas it was recurring in case of cast films of the blended mixture, and neat NC films. Altogether, the together the work details a novel path to integrate an organized lignin and cellulose molecular structure, albeit modified from their native form, into a three-dimensional composite material. / Ph. D.

This thesis presents applications of the layer-by-layer (LbL) assembly technique in fabrication of thin films with a primary focus on design and development of electrochromic devices. The optical properties of electrochromic materials change as they alter between redox states. The morphology and properties of LbL-assembled thin films can be modified by varying several processing factors such as dipping duration, ion type, ion concentration, pH, molecular weight, and ionic strength. In the present work, several factors of LbL assembly process were manipulated to tailor electrochromic thin films of desired attributes. An electrochromic device (ECD) with fast optical switching speed was designed and constructed based on poly(3,4-ethylenedioxythiophene):poly(styrenesulfonate) (PEDOT:PSS). This device exhibited optical switching speeds of 31 and 6 ms for coloration and decoloration respectively, on a 60 mm2 area. Poly(aniline 2-sulfonic acid) (PASA) is a relatively new ionic polymer, and its electrochromic properties have not been previously investigated in much detail. PASA thin film showed several redox states corresponding to color changes from dark blue to gray as it passed different redox states. One particularly interesting and promising design for ECDs is dual electrochrome. Dual electrochrome ECDs based on PANI and polyaniline (PASA) are investigated in this thesis. The PANI/PASA thin film showed superior spectroelectrochemical properties compare to other ECDs reported here or elsewhere. An electrode with single wall carbon nanotubes (SWCNTs) coating was tested as the substrate for an ECD based on poly[2-(3-thienyl) ethoxy-4-butylsulfonate] (PTEBS) to examine performance of the electrochromic polymer on a substrate other than an indium tin oxide (ITO) electrode. Compared to ITO, the SWCNT based device exhibited superior properties. / Master of Science

Electrical properties of Si- and Mg-doped InAs epitaxial layers grown by MOCVD were studied by performing magneto-transport measurements at different temperatures, from 300 K down to 1.2 K. The longitudinal magnetoresistance and Hall effect indicate a three-band system existing in n-type (p-type) InAs, which consists of the surface accumulation (inversion) layer, the bulk electron (hole) layer, and the nucleation layer. Therefore, a classical parabolic background in magnetoresistance due to multi-carrier occurs at low fields. With the magnetic field being further applied, a linear magnetoresistance caused by inhomogeneities is revealed. At liquid helium temperature, the Shubnikov-de Haas magneto-oscillations are also observed. These transport characterizations provide a means of analyzing the band structure at the InAs surface. In a set of n-type InAs epilayers with Si doped at different levels, the bulk electron density increases as the doping level increases. The increased ionized impurities lead to lower electron mobilities due to more Coulomb scatterings. For all the n-type InAs films, except the two active layers (surface and the bulk), the nucleation layer contributes to the film conductivity as well with an electron density of ~ 5 x 10¹⁷ cm⁻³ and a mobility of ~ 2000 cm²}/Vs. In a cooldown process, the electron density of each layer slightly and monotonically decreases whereas the mobility experiences a maximum from the competition between phonon scatterings and Coulomb scatterings. The phonon scattering overwhelms the Coulomb scatting at high temperatures, but declines as temperature decreases, thus the mobility increases. Around 100 K, the temperature-independent ionized impurity scattering becomes comparable with and starts exceeding the phonon scattering, as temperature further lowered, the screening effect of the Coulomb scattering is weakened because of the decreased carrier densities. As a result, the mobility starts dropping. The maximum mobility corresponds to a minimum resistance, which explains the non-trivial temperature dependence of the resistance in the cooldown history. For the p-type InAs film, the doping with Mg in the course of MOCVD growth allows us to obtain a large hole density and a low mobility at 300 K. At low temperatures, holes are frozen out, and a strong negative magnetoresistance with a dip at 0 field are observed, which is the antilocalization signal from accumulation electrons. This is a strong technique to probe the surface quantum states and derive the phase coherence length and the spin flip length of surface electrons. / Master of Science