This thesis consists of two parts. The first part describes the analysis and implementation of a finite element computer model for the general prediction of failure of wood members in bending or in combined bending and axial compression. Both instability and material strength failures are included. The program is verified using available analytical and test results. A good agreement with the results predicted by this program is observed. The second part describes a procedure for the structural reliability evaluation of a compression member assuming random loads and material variables. The program developed here for the reliability study links the finite element program and the Rackwitz-Fiessler algorithm for the calculation of the reliability index β. The gradient of the failure function, which is a necessary input to the Rackwitz-Fiessler algorithm, is computed numerically using the finite element routine. The results of the reliability study for a typical column problem are compared against the available results obtained by following the code procedures [as outlined in CAN3-086.1-M84 (1984)] for different slenderness ratios. A performance factor ɸ⍴ = 0.75, for compression members of any length is recommended in order to obtain a more accurate and consistent level of reliability in the design process. It is estimated that if this factor ɸ⍴ = 0.75 adopted in the current design practices, a level reliability index of the order of 4.0 can be achieved. / Applied Science, Faculty of / Civil Engineering, Department of / Graduate

The fracture problem due to the singular stresses arising from the sudden change of geometric properties around cracks and notches was studied both analytically and experimentally. The failure models of the cracked and the notched specimens were derived by using linear elastic fracture mechanics methodology, which led to the determination of the critical stress intensity factors. Experiments were conducted to determine fracture toughness for different modes as well as the effect of variations in the crack-front width, specimen size and moisture content. Subsequently, failure surfaces for cracks and notches were developed based on the experiments undertaken, describing in each case the interaction between mode I and mode II fracture toughness. To verify the reliability of these experiments, the results obtained were compared with the published literature. As an application, design curves for a 90 degrees-cracked beam and a 90 degrees-notched beam are presented. These curves allow the prediction of the failure loads due to the rapid crack propagation under different loading conditions. / Applied Science, Faculty of / Civil Engineering, Department of / Graduate

M.Tech. (Civil Engineering) / Every year, millions of Rands are being spent in rehabilitation, repairs and maintenance of reinforced concrete structures projects around the country due to corrosion. There are a number of studies and investigations that have been done recently to address the effect of corrosion on reinforced concrete structures. Concrete provides an ideal environment for steel, supplying both physical and chemical protection from corrosive attack. The effect of corrosion on reinforced concrete structures contributes to the reduction of steel cross sectional area, weakening the bond strength between steel and concrete and thereby reducing the ductility, deflection capacities and load carrying capacity of the structure. In this research, five series of three samples each of reinforced concrete beams were fabricated, some with reduced cross-sectional area and with exposed bars to simulate loss of bonding through spalling. The structural performance of these beams was tested in terms of maximum load carrying capacity, deflections and ductility ratio. The main conclusions are as follows:  As little as 1 % loss in mass of tension steel resulted in a load carrying capacity decrease of 6.9 %  Beam deflection increased by a factor of 1.5 times when the steel mass loss level was 14%  5 % steel loss in mass resulted in a 16.5 % decrease in relative ductility.

A radio-frequency positive-ion source and a fifty-kilovolt linear accelerator were designed and constructed in order to produce sizable quantities of hydrogen, helium, nitrogen, neon, and argon. Plans were then made to equip this ion source with charge-exchange apparatus suitable for charge-exchange cross-sectional measurements. It is the purpose of this paper to present the design of the equipment and to present operational knowledge of the equipment and of ion beams which are producible.

In the present study, an analysis of plain and reinforced concrete beams under monotonic loading was made based on the Fixed Smeared Crack approach. The objectives of this research were to analyze the nonlinear behavior of the selected cases of analysis and to propose an alternative and simple model for the analysis of beams under service loadings, by means of Committee 435 of the American Concrete Institute. A brittle model for concrete and a linear-elastic behavior for steel reinforcement bars were considered. Results are presented through force-displacement curves and the sequence of cracking propagation. Also, a comparison of calculated instantaneous deflections of simply supported beams was made between the proposed model and other researches. It was verified that the proposed algorithm can predict adequately the cracking process and the deflections of beams subjected to service loadings, taking into account experimental results from other authors.

Printed lens antennas offer the possibility of realizing shaped beam patterns using no more complexity than is required for pencil beam patterns. Shaped beam patterns can be obtained by appropriately determining the complex transmission coefficient required for each cell (or element) of the printed lens, taking into account the varying feed field over the input surface of the lens. Certain ranges of transmission coefficient amplitude and phase are undesirable (eg. too low an amplitude implies a large reflection at the lens input surface). It would be preferable to constrain the range of values that the transmission coefficient can take as an integral part of the lens synthesis procedure, and thus the transmission coefficient itself needs to be the synthesis variable. In this thesis a synthesis technique for doing this is developed based on the method of generalized projections, modified to “operate” in the space of transmission coefficients. This makes it possible to immediately perceive what influence constraints on the actual transmission coefficients have on the possible radiation pattern performance. In addition, an approach that allows one to constrain the transmission coefficient to values that must be selected from an available database of transmission coefficients is incorporated into the synthesis technique.

This work is done as a small facet of a much larger study on efficient control of indoor air environments. Halton passive chilled beams are used to cool rooms and the focus of this work is to model the beams. This work also reviews the mesh making process in Gmsh. ANSYS Fluent was used throughout the entire research and this thesis describes the software and a careful description of the case study. / Master of Science

The energy distribution in an electron beam transmitted by a thin aluminum film was studied by a retarding potential method in which the energy analysis could be made as a function of the polar angle and azimuth of scattering. The results indicated that the transmitted beam contained electrons which had lost essentially zero energy and groups of electrons which had suffered losses of approximately15 ev and 25 ev and possib1y 75 ev. In addition there appeared to be a continuous background of energy losses over the range studied (from 0 to 80 ev). The resolution of the energy analyser was limited by the separation of the data points and was not adequate to observe the dependence of the 15 ev energy loss (plasma loss) on the polar angle of scattering. This study demonstrated the feasibility of this method of energy analysis as well as the short-comings of the apparatus. Suggestions for future improvements are made.

A recurring theme in my research is that mathematical matrix methods may be used in a wide variety of physics and engineering applications. Transfer matrix techniques are conceptually and mathematically simple, and they encourage a systems approach. Once one is familiar with one transfer matrix method, it is straightforward to learn another, even if it is from a completely different branch of science. Thus it is useful to overview these methods, and this has been done here. Of special interest are the applications of these methods to laser optics, and matrix theorems concerning multipass optical systems and periodic optical systems have been generalized here to include, for example, the effect of misalignment on the performance of an optical system. In addition, a transfer matrix technique known as generalized beam method has been derived to treat misalignment effects in complex optical systems. Previous theories used numerical or ad hoc analytical solutions to a complicated diffraction integral. The generalized beam matrix formalism was also extended to higher-order beam modes of lasers and used to study mode discrimination in lasers with misaligned complex optical elements.

No description available.