In this thesis, the moisture characteristics of spruce plywood are studied experimentally and numerically with special attention given to moisture storage and release as the indoor humidity changes diurnally. This is referred to as the moisture buffering capacity. Two test facilities (a glass jar facility and a transient moisture transfer facility) are used to measure the moisture accumulation and temperature and relative humidity profiles within spruce plywood. These measured data are used to determine the moisture buffering capacity of spruce plywood and validate a one-dimensional transient numerical model that can be used to calculate the transient heat and moisture transfer in spruce plywood. There is good agreement between the measured and simulated results over the range of test variables investigated. </p><p>This validated numerical model is used to investigate the effect of initial conditions, boundary conditions, thickness and humidity step change on the moisture buffering capacity. In addition, sensitivity studies are performed to investigate the effect of variations in material properties used in the numerical model. The properties that are considered in these sensitivity studies are the sorption isotherm, effective thermal conductivity, heat of sorption and effective diffusion coefficient. These studies show that the sorption isotherm has the greatest effect on the moisture buffering capacity, as well as the temperature and relative humidity profiles within spruce plywood. For example, a ± 10% change in sorption isotherm has a ± 7%, ± 6% and ± 10% effect on the moisture buffering capacity, and the relative temperature and relative humidity change, respectively. </p><p>This thesis also verifies the moisture diffusivity property for spruce plywood, which was developed by Olutimayin and Simonson (2005) to account for moisture storage in cellulose insulation for a single step change in humidity. It was found that for spruce plywood, the moisture penetration depth may be over predicted by an order of magnitude when moisture storage is neglected using a transient analytic solution which does not include moisture storage.

<p>Ultrarelativistic heavy-ion collisions at the Relativistic Heavy-Ion Collider (RHIC) are thought to have produced a state of matter called the Quark-Gluon-Plasma (QGP). The QGP forms when nuclear matter governed by Quantum Chromodynamics (QCD) reaches a temperature and baryochemical potential necessary to achieve the transition of hadrons (bound states of quarks and gluons) to {it deconfined} quarks and gluons. Such conditions have been achieved at RHIC, and the resulting QGP created exhibits properties of a near perfect fluid. In particular, strong evidence shows that the QGP exhibits a very small shear viscosity to entropy density ratio &eta/s, near the lower bound predicted for that quantity by Anti-deSitter space/Conformal Field Theory (AdS/CFT) methods of &eta/s = $hbar$/ 4 &pi $k_B$ where $hbar$ is Planck's constant and $k_B$ is Boltzmann's constant. As the produced matter expands and cools, it evolves through a phase described by a hadron gas with rapidly increasing $eta/s$.</p><p>This thesis presents robust calculations of $eta/s$ for hadronic and partonic media as a function of temperature using the Green-Kubo formalism. An analysis is performed for the behavior of $eta/s$ to mimic situations of the hadronic media at RHIC evolving out of chemical equilibrium, and systematic uncertainties are assessed for our method. In addition, preliminary results are presented for the bulk viscosity to entropy density ratio $zeta/s$, whose behavior is not well-known in a relativistic heavy ion collisions. The diffusion coefficient for baryon number is investigated, and an algorithm is presented to improve upon the previous work of investigation of heavy quark diffusion in a thermal QGP. </p><p>By combining the results of my investigations for $eta/s$ from our microscopic transport models with what is currently known from the experimental results on elliptic flow from RHIC, I find that the trajectory of $eta/s$ in a heavy ion collision has a rich structure, especially near the deconfinement transition temperature $T_c$. I have helped quantify the viscous hadronic effects to enable investigators to constrain the value of $eta/s$ for the QGP created at RHIC.</p> / Dissertation

We study some topological properties of dynamical systems. In particular the rela- tionship between spatio-temporal chaotic and Li-Yorke sensitive dynamical systems establishing that for minimal dynamical systems those properties are equivalent. In the same direction we show that being a Li-Yorke sensitive dynamical system implies that the system is also Li-Yorke chaotic. On the other hand we survey the possibility of lifting some topological properties from a given dynamical system (Y, S) to an- other (X, T). After studying some basic facts about topological dynamical systems, we move to the particular case of interval maps. We know that through the knowl- edge of interval maps, f : I → I, precious information about the chaotic behavior of general nonlinear dynamical systems can be obtained. It is also well known that the analysis of the spectrum of time series encloses important material related to the signal itself. In this work we look for possible connections between chaotic dynamical systems and the behavior of its Fourier coefficients. We have found that a natural bridge between these two concepts is given by the total variation of a function and its connection with the topological entropy associated to the n-th iteration, fn(x), of the map. Working in a natural way using the Sobolev spaces Wp,q(I) we show how the Fourier coefficients are related to the chaoticity of interval maps.

Reservoir fluids are rarely found in homogeneous structures having homogeneous properties. The various elements and processes of the petroleum system result in complex fluid distributions and compositions. A sound understanding of these complexities can avoid disappointing results and costly mistakes when designing the completion and production of the reservoir. The earlier these complexities are understood in the exploration phase, the better are the chances of a successful decision making process in the design phase of the project. Assessing reservoir compartment connectivity is of paramount importance for a optimal field development. Recent technological advances in wireline formation testing and sampling provide asset teams with a new methodology to evaluate in situ fluid properties and reservoir connectivity. After a review of the technology of downhole fluid analysis (DFA), the currently available methods of modeling equilibrated fluid gradients are presented. Fluid composition equilibrium is a stationary state where all components have reached zero mass flux. A reservoir model is designed to simulate numerically equilibration processes over geologic timescales at isothermal conditions where diffusion and gravity are the active mechanisms. A variety of initial conditions and reservoir fluid types is considered. Non-equilibrium fluid gradients and their transient behavior as they evolve towards fluid composition equilibrium are the main interest of this study. The results are compared in case studies, that are available in published literature. The modeling methods allow modeling of vertical and lateral fluid gradients. After a discussion of the cases, this thesis gives recommendations on 1) what fluid properties should be assessed and 2) how many data points are needed to reduce the chance of misinterpretation of non-equilibrium gradients in the presence of faults. To make best use DFA data, the property that exhibits the largest gradient needs to be investigated, as it yields the greatest potential to assess connectivity. The shape of the distribution of fluid composition within a compartment is found to be an important part in investigating reservoir connectivity. During data acquisition efforts should be made to acquire enough data points to reveal this shape. In combination with the presented techniques to identify non-equilibrium conditions, this will optimize DFA data acquisition and maximize the value of the data.

A Mandarin speech recognition system for addresses based on MFCC, hidden Markov model (HMM) and Viterbi algorithm is proposed in this thesis. HMM is a doubly stochastic process describing the ways of pronunciation by recording the state transitions according to the time-varing properties of the speech signal. In order to simplify the system design and reduce the computational cost, the mono-syllable structure information in Mandarin is used by incorporating both mono-syllable recognizor and HMM for our system. For the speaker-dependent case, Mandarin address inputting can be accomplished within 60 seconds and 98% correct identification rate can be achieved in the laboratory environment.

In the beginning of this study, Molecular dynamics simulation is utilized to investigate the behavior of water molecules confined between two Au plates of (001) planes separated by gaps of 24.48, 16.32, 12.24, 11.22, and 10.20 . The simulation results indicate that the arrangements of the water molecules are dependent on the gap size. An inspection of the variation of the self-diffusion coefficients with the gap size suggests that the difference between the dynamic properties of the water molecules in the z-direction and the x-y plane decreases as the distance between the two Au plates increases. Moreover, we discuss the effects of different lattice structures, (100), (110) and (111)¡Aon the water molecules. The simulation results indicate that the arrangements of the water molecules are dependent on Au plate surface structures. The adsorption of the plate creates flat water layers in the proximity of each plate surface for (100) and (111) cases, but wave-like water layer for Au (110) plate. The absorbed water layer is the most close to plate surface for (110) lattice structure. Moreover, the self-diffusion coefficient in the z-direction for (110) case is the largest, meanwhile, the water molecules have a greater ability to diffuse in the x-y plane for (100) case. Finally¡Athe density distribution, velocity profile, and diffusion coefficients of the water film in a Couette flow are studied. Shear viscosity and its dependence on the shear rate of the water film are also examined in the present research. The diffusion of the whole film increases dramatically as the shear rate greater than a critical value. The shear viscosity decreases as the shear rate increases, especially for the water film with a small thickness, which implies the shear-thinning behavior for viscosity of the nanoconfined film. Moreover, increase in shear viscosity with a decrease in the film thickness can also be found in the present study.

A system of Chinese resume by speech construction is developed by the use of a novel segmentation mechanism and the classical Hidden Markov Model. The recognition system is based on both mono-syllable HMM's and speech-text alignment schemes. Experimental results indicate that the amount of training materials used for feature extraction can be greatly reduced, and the text content of the recorded speech training data can be different from those of the recognition tasks as well. Each phrase in the resume can be identified within one second, that is approximately the same as the graduate did last year. Furthermore, the user interface of the resume system has been redesigned and polished by the GTK toolkit in order to enable event-driven X-window operations.

A design of speech recognition system for Chinese names of historical figures around the world is proposed in this thesis. A speech database of approximately forty-six thousand Chinese names is collected and recorded twice for system evaluation. This system applies Mel-frequency cepstrum coefficients, monosyllable HMM¡¦s and speech-text alignment scheme to accomplish initial candidate selection. A Mandarin pitch identification mechanism is then followed to increase the correct rate and obtain the final answer. The experimental results indicate that a 90% correct identification rate can be achieved, under the condition that the first session recording material is used for training and the second one for testing. For the speaker dependent case, the correct name can be recognized within 1.5 seconds, using a PC with an Intel Celeron 2.4 GHz CPU and RedHat Linux 9.0 Operation System.

A four-session text independent, TV-recorded audio-video database for speaker recognition is collected in this thesis. The speaker data is used to verify the applicability of a design methodology based on Mel-frequency cepstrum coefficients and Gaussian mixture model. Both single-session and multi-session problems are discussed in the thesis. Experimental results indicate that 90% correct rate can be achieved for a single-session 3000-speaker corpus while only 67% correct rate can be obtained for a two-session 800-speaker dataset. The performance of a multi-session speaker recognition system is greatly reduced due to the variability incurred in the recording environment, speakers¡¦ recording mood and other unknown factors. How to increase the system performance under multi-session conditions becomes a challenging task in the future. And the establishment of such a multi-session large-scale speaker database does indeed play an indispensable role in this task.

This paper presents text-independent speaker identification results for medium-size speaker population sizes up to 400 speakers for TV speech and TIMIT database . A system based on Gaussian mixture speaker models is used for speaker identification, and experiments are conducted on the TV database and TIMIT database. The TV-Database results show medium-size population performance under TV conditions. These are believed to be the first speaker identification experiments on the complete 400 speaker TV databases and the largest text-independent speaker identification task reported to date. Identification accuracies of 94.5% on the TV databases, respectively and 98.5% on the TIMIT database .