No description available.

The structure of a stable smooth wall zero pressure gradient turbulent boundary layer is investigated experimentally in order to determine the dominant outer region structure and to develop a hypothetical generalized boundary layer flow model. Three hot wire configurations, two vertically separated X-wires and a leading straight wire, a horizontal rake of 5 straight wires, and a vertical rake of 5 straight wires were used in the experiments, conducted at Reɵ = 8200. The basis for data reduction procedures came from crosscorrelations and the Variable Interval Time Average (VITA) technique. Three structure types are reported in the literature to be important: streaks and counter rotating streamwise vorticity, wall scaled hairpins or ring vortices, and large scale (0(ઠ)) bulges. A simple pictorial model consisting of three Reɵ dependent interdeveloping stages, which integrate all three structure types, is presented and discussed in relation to the literature and experiments performed. The rake data indicate that the positive ([formula omitted]u/[formula omitted]t) VITA detected velocity front has a scale much larger than that of the wall scaled eddies which typically have a scale of 100-300 y[formula omitted], and that this velocity front exhibits characteristics that are consistent with the trailing velocity front described in the model. The general convection velocity from basic crosscorrelations and the convection velocity of the positive VITA detected velocity front both had values 90-100% of the local mean velocity over most of the boundary layer. Evidence of small scale structure concentration on the downstream edge of the trailing velocity front is presented. A new method used to determine the average structure inclination angle associated with the trailing velocity front is presented and demonstrates that the generalized structure inclination angle, calculated from basic crosscorrelations between vertically separated sensors, does not indicate structure shape, but is associated with the bulk flow associated with the structure. The new method appears to give results that are consistent with flow visualization and more accurately estimates the inclination angle associated with the most dominant feature of the outer flow, the positive VITA velocity front. Although the model presented is somewhat crude and further development and refinement are required, the model appears to agree with most data in the literature, as well as the present experimental results. / Applied Science, Faculty of / Mechanical Engineering, Department of / Graduate

Analytical and numerical analyses along with experimental results are presented for laminar forced convection from a heated flat plate. The heating is of either the constant flux or constant temperature type with a discontinuity in the lateral as well as the free-stream direction. The principal objective was to examine the disturbance created by the lateral discontinuity in heating. Analytical solutions of the boundary layer type relating the wall temperature variation to a specific heat flux distribution were found. These involve Heaviside calculus, Fourier transform techniques and the Lévque velocity approximation, and although formally correct only as the Prandtl number approaches infinity, are found to be fairly accurate for Prandtl numbers as low as unity. Reduction of the energy equation to a two-dimensional elliptic equation by a double similarity transform enabled the calculation of numerical solutions valid in any free-stream cross-section removed from the leading edge. Results for technologically interesting cases of heated strips were found by superposition. The error in assuming such strips to be two-dimensional throughout their lateral extent is calculated. A velocity profile corresponding with the Blasius profile and a cross-sectional temperature field above a wind tunnel model of a heated flat plate were measured with a constant temperature anemometer and a rake of miniature thermocouples respectively. A temperature profile well into the heated portion is of two-dimensional form and the temperature field at the lateral step correlates with a numerical solution. In general, there is good agreement between all three approaches. / Applied Science, Faculty of / Mechanical Engineering, Department of / Graduate

Some conductive and convective heat transfer problems with radiative boundary conditions are analysed theoretically. Three specific problems have been analysed. The study has, therefore, been divided into three parts. In Part I the temperature distribution produced in-long, solid circular and rectangular cylinders and a solid sphere in interplanetary space is studied. The solid bodies receive parallel radiation flux on one side and emanate radiant energy to their surroundings at zero degree Rankine. Steady state, constant physical and surface properties, and no heat loss by convection are assumed. Solution of the linear conduction equation with nonlinear boundary conditions is obtained by two approximate, semi-analytical methods, (i) point matching and (ii) least-squares fitting. The results are compared with earlier results obtained by a variational method. The least-squares fit method appears to be most suitable regarding accuracy and simplicity in computation. Its accuracy does not appear to depend appreciably either on the radiation-conduction parameter or on the surface absorptivity. The effect of semi-grayness of the receiving surface is analysed. In Part II the heat transfer characteristics of a circular fin dissipating heat from its surface by convection and radiation are analysed. The temperature is assumed uniform along the base of the fin and constant physical and surface properties are assumed. There is radiant interaction between the fin and its base. Two separate situations are considered. In the first situation heat transfer from the end of the fin is neglected. Solution of the linear conduction equation with nonlinear boundary conditions has been obtained by the least-squares fit method. A solution has also been obtained by the finite difference method and the results compared. Results are presented for a wide range of environmental conditions and physical and surface properties of the fin. In the second situation heat transfer from the end of the fin is also included in the analysis. The solution is obtained by a finite difference procedure. It is shown that neglecting heat transfer from the end is a good approximation for long fins or for fins of high thermal conductivity material. In Part III the problem of laminar heat transfer in a circular tube under radiant heat flux boundary conditions has been analysed. Fully developed velocity profile is assumed and the tube is considered stationary. A steady radiant energy flux is being incident on one half of the tube circumference while the fluid emanates heat through the wall on all sides by radiation to a zero degree temperature environment. A solution by finite difference procedure has been obtained. The temperature distribution and the Nusselt number variation are presented for a wide range of the governing physical parameters. / Applied Science, Faculty of / Mechanical Engineering, Department of / Graduate

M.Ing. / The goal of the project was to stimulate aspects of the atmospheric boundary layer. As a result of the complexity of the derived models, it was decided to rather do an experiment on turbulent Couette flow to simulate. some aspects of the atmospheric boundary layer. A Couette flow channel with a maximum wall speed of 1 m/s was built. Experiments were done at two Reynolds numbers, l600 and 2500, corresponding to a channel height of 0.06 m and belt speeds of 0.4 mI and 0.58 m/s, respectively. The Coutte channel has the facility that the supper stationary wall can be heated or cooled. Experiments were carried out with positive and negative temperature gradients for both the Reynolds numbers giving the following results. By heating the top wall a stratified flow results which leads to no mixing of the flow properties. This phenomena is analogous to night time conditions in the atmosphere, when layers at different temperatures form a layered structure with little mixing between layers.

This thesis uses the Immersed Boundary Method (IBM) to simulate the movement of a human heart. The IBM was developed by Charles Peskin in the 70’s to solve Fluid-Structure Interaction models (FSI). The heart is embedded inside a fluid (blood) which moves according to the Navier-Stokes equation. The Navier-Stokes equation is solved by the Spectral Method. Forces on the heart muscle can be divided into two kinds: Active Force and Passive Force. Passive includes the effect of curvature (Peskin’s model), spring model, and the torsional spring (or beam) model. Active force is modeled by the 3-element Hill model, which was used in the 30’s to model skeletal muscle. We performed simulations with different combinations of these four forces. Numerical simulations are performed using MATLAB. We downloaded Peskin’s code from the Internet and modified the Force.m file to include the above four forces. This thesis only considers heart muscle movement in the organ (macroscopic) scale.

State-of-the-art deep learning methods for outlier detection make the assumption that outliers will appear far away from inlier data in the latent space produced by distribution mapping deep networks. However, this assumption fails in practice,because the divergence penalty adopted for this purpose encourages mapping outliers into the same high-probability regions as inliers. To overcome this shortcoming,we introduce a novel deep learning outlier detection method, called Outlier Preserving Distribution Mapping Autoencoder (OP-DMA), which succeeds to map outliers to low probability regions in the latent space of an autoencoder. For this we leverage the insight that outliers are likely to have a higher reconstruction error than inliers. We thus achieve outlier-preserving distribution mapping through weighting the reconstruction error of individual points by the value of a multivariate Gaussian probability density function evaluated at those points. This weighting implies that outliers will result in an overall penalty if they are mapped to low-probability regions. We show that if the global minimum of our newly proposed loss function is achieved,then our OP-DMA maps inliers to regions with a Mahalanobis distance less than \delta, and outliers to regions past this \delta, \delta being the inverse ChiSquared CDF evaluated at 1−\alpha with \alpha the percentage of outliers in the dataset. We evaluated OP-DMA on 11 benchmark real-world datasets and compared its performance against 7 different state-of-the-art outlier detection methods, including ALOCC and MO-GAAL. Our experiments show that OP-DMA outperforms the state-of-the-art methods on 7 of the datasets, and performs second best on 3 of the remaining 4 datasets, while no other method won on more than 1 dataset.

No description available.

No description available.

No description available.