<p>A computer model is developed for the solution of regional network systems on a minimum cost basis. Different network problem types are identified to define the scope of the model. Generally, the problem involves the determination of the optimum conveyance schedule required to supply a set of consumer nodes with a commodity or public service provided by one or more processing centres (i.e. water supply, solid waste or wastewater collection networks). The system costs include costs incurred due to conveyance of material and the processing of that material; both of these cost components exhibit economies of scale and generally lead to the centralization of processing.</p> <p>A mathematical statement of the problem is developed which is applicable to all network types, can be utilized easily and efficiently in a digital computer and facilitates the use of a variety of optimization routines. A number of algorithms, ranging from linear approximation to nonlinear gradient search routines, are investigated for possible inclusion in the model with the advantages and disadvantages of each being identified.</p> <p>A modular package, NETSOL, is developed which facilitates the use of alternate optimization routines and allows for the inclusion of complex design functions in the computation of system costs. An interactive command structure permits the user to modify the network system parameters thereby combining intuitive design with the capacity to select an optimal solution from a large number of alternatives. Thus a model is obtained which is flexible enough to answer many of the questions that arise in network problems and also determine the optimum.</p> <p>In view of the disadvantages found in the existing optimization techniques (e.g. convergence to local minima), the properties of the network problems are investigated in detail to isolate any special characteristics. For a system involving separable concave cost functions, the minimum cost solution to a network problem lies at one of the vertices formed by the problem constraints. A new solution algorithm, HYVRST, is developed which takes advantages of this important property utilizing a direct search technique. This results in an efficient and stable algorithm with good convergence properties. A number of example are presented to test the optimization algorithms and demonstrate the usefulness of the NETSOL package in solving regional network problems.</p> / Master of Engineering (ME)

<p>A 2 inch diameter, conical-shaped glass hydrocyclone, operating without an air core, was used in this study. The geometric dimensions of the cyclone followed closely with the optimum design conditions determined for solid/liquid systems by Rietema (R-1) and used by Burrill and Woods (B-3) for liquid/liquid systems. Distilled water was used as the continuous phase. Dispersed oil phases studied include: butanol, methyl isobutyl ketone (MIBK) , toluene and kerosene.</p> <p>For each of the liquid/liquid systems, the efficiency of separation was determined as a function of volume split, oil/water phase ratio and feed flowrate. Differentiation of the liquid/liquid systems, in terms of physical properties, was based primarily on interfacial tension. Density difference and viscosity of the dispersed phase, were comparable from one system to another. Mixing energy used to disperse the oil phase in the water phase, geometric dimensions and temperature were constant throughout the work. The range of the operating variables were as follows:</p> <p>i) oil/water phase ratio 0.160 to 1.00</p> <p>ii) feed flowrate 100 to 365 mL/s</p> <p>iii) interfacial tension 2.0 to 30.0 mN/m</p> <p>iv) volume split 0.17 to 3.90</p> <p>For each system studied, photographs were taken at the inlet and outlets leading to and from the cyclone, respectively, to determine the drop size of the dispersed phase.</p> <p>The second part of the present work considered the influence that the mixing energy had on the effects of oil/water ratio and feed flowrate as studied in the first part.</p> <p>The efficiency of separation (Es ) is defined as follows:</p> <p>[equation removed]</p> <p>where Y and Q represent volume fraction of light phase and flowrate, respectively, while the subscripts denote specific orifice location on the hydrocyclone.</p> <p>From the first part the efficiency of separation in the cyclone was a very important function of volume split. The effect of the feed flowrate on separation in the cyclone was dependent on the interfacial tension. The effect of oil/water ratio was dependent on the rate of coalescence. Based on the photographic work, coalescence occurred in the cyclone for several of the systems studied.</p> <p>The majority of past work has varied the mixing energy with a change in the feed flowrate. As a result, drop size varied. Present work revealed that this reversed the effect that feed flowrate had on the separation in the cyclone with mixing energy constant.</p> <p>It was not possible to obtain two pure phases from the hydrocyclone for any of the systems studied. One pure phase, however, was achieved for three of the four systems studied. A relatively pure water phase (≥ 99%) was obtained at the underflow for the MIBK/water, toluene/water and kerosene/ water systems. The highest values of the optimum Es were 67, 57 and 62%, respectively, for each of these systems. For toluene/water and kerosene/water systems, the interfacial tension was sufficiently high to prevent significant drop breakup when the feed flowrate was increased. Under similar circumstances, drop breakup predominated for MIBK/water and butanol/water systems due to the lower values of interfacial tension. With butanol/water, a significant amount of light phase was found in the underflow. The optimum Es for this system was only 26%.</p> <p>It was noted that the efficiency of separation, Es, increased sharply at first and then decreased gradually with increasing volume split. The optimum volume split occurred at a value greater than the feed phase ratio for all systems studied. The optimum volume split occurred at a value ranged from 100 to 500% greater than the feed phase ratio. Since complete coalescence does not occur inside the cyclone, it is not possible to have the optimum split equivalent to the feed phase ratio. Continuous phase trapped in the interstides results in the optimum split equivalent to a value greater than the feed phase ratio. A simple mass balance model was used to describe the effects of volume split. From this model the interstitial volume could be inferred for all conditions. Combining this information with models for breakup, coalescence and hindered setting yielded a semi-quantitative explanation of all the trends observed.</p> <p>The feasibility of using the hydrocyclone to separate emulsions is based on achieving at least one pure phase. If this achievement is accomplished, then it is possible to reduce the volume requirement of a gravity settler. The role of the hydrocyclone is basically one of a preliminary stage in the physical separation process. If, on the other hand, it is not possible to have one pure phase, no useful purpose is served by the cyclone. Consequently, the butanol/water system can not be considered a feasible system to be separated by a hydrocyclone.</p> / Master of Engineering (ME)

<p>A theoretically sound and practical method of flexible pavement design remains one of the most needed, though widely unappreciated, aspects of transportation engineering. Papers presented at the Third International Conference on the Structural Design of Asphalt Pavements in 1972 greatly increased the awareness of design engineers of the testing problems involved in this field. The generated interest in rational, as opposed to empirical, flexible pavement design approaches brought about a pressing need for material characteristics for use in available theoretical modes.</p> <p>This study involved the development of laboratory equipment capable of providing values of resilient modulus, MR, and Poisson's ratio, v, of asphalitic concrete specimens. These properties are required as input for elastic layer analyses of flexible pavement systems. It was recognized that the developed equipement must be capable of providing reliable results at similar costs to conventional Marshall or Hveem tests, and should be adaptable to realistically simulate site conditions. It is considered that the equipment described herein satisfies these conditions.</p> <p>Test results on polymeric calibration samples compare favourably to those obtained during previous work on the resilient modulus of asphalt mixes. The testing program identified temperature, asphalt content and aggregate type as parameters that affect the resilient modulus. A future phase of this study onvolves a more comprehensice testing program in these areas, in addition to studying the effects of confining pressure and stress levels on material properties using equipment suggested from this initial testing program.</p> <p>Comparative design studies using available theoretical methods and generated stiffness parameters indicate possible cost savings as a result of thickness requirements for pavement systems comprised of mixes with higher MR values. Conventional empirical approaches do not permit similar thickness reductions for higher quality (strength) mixes.</p> / Master of Engineering (ME)

<p>This thesis deals with the estimation of wave directional spectra and applications to the study of surface gravity water waves. Theoretical foundations and testing procedures are established to evaluate and compare different methods of extracting the wave directional spectrum from a wave record. An integrated software package for the analysis of directional seas is developed and used to test all methods, and to identify their properties, characteristics and biases. As a result, guidelines for the use of these methods are drawn. The findings indicate that all of the methods currently used for wave directional spectrum estimation have drawbacks. As a result, three new methods are proposed and tested against current methods. Test results indicate that a proposed closed-form of the Maximum Likelihood Method is the best choice, from both a theoretical and computational point of view. The new method was shown to outperform all other estimates for both heave-pitch-roll and wavestaff data. Field data from the Atlantic Ocean, Lake Ontario and Lake St. Clair (spanning two orders of magnitude in size) is investigated in an attempt to demonstrate the resolution potential of the newly developed method of estimating the wave directional spectra. As a result, completely decoupled spectra are observed for the first time in rapidly turning winds, and a clear relationship between the wave relaxation parameter and the wave age is established. Directional spreading parameter values are found to be higher than the established values in previous studies, for both Lake Ontario and Atlantic Ocean data. In addition, the Atlantic Ocean directional spectra are found to be narrower than their Lake Ontario counterpart. Finally, wave measurements in Lake St. Clair indicate that the structure of directional spectra can be very complicated, even in small lakes, and that a strong shear current can cause not only the refraction of an incoming wavefield, but can also inhibit the generation of waves propagating directly against it.</p> / Doctor of Philosophy (PhD)

<p>ABSTRACT</p> <p>The first application of the tuned liquid damper (TLD) to mitigate the</p> <p>dynamic vibrations of structures was only around 20 years ago and has just been</p> <p>recently applied in North America. TLDs are partially fluid filled tanks (usually</p> <p>water) with a fundamental sloshing frequency tuned close to the frequency of the</p> <p>dynamic mode of structural vibration to be suppressed. Water alone is</p> <p>insufficient to achieve the level of damping typically required for design.</p> <p>Damping devices are often submerged in the water to greatly increase the inherent</p> <p>TLD damping. The damping device investigated in this study is a thin sharpedged</p> <p>horizontal-slat screen. TLDs with such screens of a particular solidity are</p> <p>designed for one target amplitude of structural response and have limited</p> <p>efficiency over a range of structural response. To increase the efficiency, the</p> <p>concept of smart screens is introduced in this study.</p> <p>Smart screens is the name given to a damping screen that alters its fluid</p> <p>pressure-loss characteristics at differing levels of excitation, (ideally) in a passive</p> <p>state of control. Symmetric fixed-angle screens and oscillating (rotating) parallellinked</p> <p>screens are experimentally investigated inside a rectangular TLD on a</p> <p>shake-table under sinusoidal motion in this study.</p> <p>TLDs have similar principles to common tuned mass dampers (TMD) and</p> <p>are analyzed accordingly. The TLD equipped with fixed-angle screens is</p> <p>modelled with linear numerical fluid models to simulate the TLD performance for preliminary design purposes. An inclined screen alters the pressure-loss</p> <p>characteristics from its typical vertical position, which in turn changes the</p> <p>inherent TLD damping, allowing damping to be controlled by simple screen</p> <p>rotation. The analytical models, including the utilization of a pressure-loss</p> <p>coefficient for an inclined horizontal-slat screen in oscillatory flow developed in</p> <p>this study, are compared with experimental results to verify their accuracy and</p> <p>ascertain limitations.</p> <p>Oscillating smart screens are investigated mainly for their practical</p> <p>consideration in a preferred passive mode of control. The screens rotate</p> <p>automatically with changes in fluid velocity (or excitation amplitude). Their</p> <p>ability to maintain a near-constant amount of TLD damping (or resonant energy</p> <p>dissipation) is examined. Other implementations of (passive) smart screens are</p> <p>possible and suggestions for future study are recommended.</p> <p>A TLD equipped with the mathematically modelled symmetric fixed-angle</p> <p>screens is theoretically investigated in a hypothetical structure-TLD system. This</p> <p>system demonstrates the ability of a smart screen to change its damping</p> <p>characteristics-altering the angle of inclination in this study-over a range of</p> <p>structural response thereby maintaining an optimal level of efficiency over a</p> <p>range of structural response accelerations.</p> / Master of Applied Science (MASc)

<p>The accuracy of the δ-MnO₂ method in determining Conditional Stability Constants (CSC's) for trace metal - organic complexes depends on the precise modeling of trace metal uptake by δ-MnO₂. Accordingly, characteristics of δ-MnO₂ and its adsorption behaviour for Cu were studied.</p> <p>Laboratory preparation of δ-MnO₂ is influenced by pH. A neutral redox process and an acid reduction of Mn⁺⁷ yield δ-MnO₂, but an alkaline oxidation of Mn²⁺ does not produce δ-MnO₂. Positive identification of MnOᵪ as δ-MnO₂ includes confirmation of characteristics such as x > 1.9, an adsorption capacity (ɼmax) ≥ 0.25 mol Cu/mol MnO₂, an amorphous XRD pattern, a specific TEM morphology and 1 < pHzpc < 2.</p> <p>Natural and simulated aging of δ-MnO₂ depends on temperature, pH and K content of the solid, and causes reduction of surface area and adsorption capacity by a factor 1.5 (natural) to 7 (simulated).</p> <p>Adsorption of trace metals onto δ-MnO₂ is described almost exclusively by a linearized Single Langmuir isotherm in the literature. This study found that Cu adsorption on δ-MnO₂ deviates from Langmuir linearity at low surface coverages. Explanations include the existence of two adsorption sites (Double Langmuir model) and non-constant activity of surface groups. The Implicit Langmuir expression is derived from surface complexation theory to model the latter, and has the form:</p> <p>[equation removed]</p> <p>where ɼads (= mol Cu adsorbed/mol MnO₂) and the free Cu²⁺ concentration in solution are measured. r ,B and n are the adsorption parameters. This model predicts the observed deviation from linearity. The bindings energy depends on the pH, the H⁺/Cu²⁺ exchange ratio (n), the surface coverage (ɼads/ɼmax) and a constant B.</p> <p>Adsorption isotherms were obtained over a pH range of 6 to 8.5, and a Cutotal range of 1 - 40 μM with glycine added to prevent precipitation of Cu. Adsorption results were fitted to Single, Double and Implicit Langmuir models, and compared. The Implicit Langmuir model describes Cu uptake by δ-MnO₂ most accurately, and CSC's for Cu-NTA and Cu-glycine complexes were readily determined using this model.</p> <p>¹⁴C studies indicated that NTA, glycine and aspartic acid do not adsorb on δ-MnO₂.</p> / Doctor of Philosophy (PhD)

<p>The free vibrations of skirt supported pressure vessels are studied in this thesis; both cantilevered and fixed-pinned systems are considered. A hierarchy of models, ranging from a rigid mass (vessel) supported by massless Euler-Bernoulli beam (skirt) to a model in which both components are represented by Timoshenko beams, is subjected to analysis. Several typical numerical examples are considered for both sets of boundary conditions. The results of these calculations indicate that whereas the cantilevered system may be modeled with fair accuracy, compared to the most sophisticated model considered, by a rigid mass supported by a massless beam capable of undergoing shear deformation, it is necessary to model all components of the fixed-pinned system by Timsoshenko beams, i.e. the most sophisticated model considered. The first two mode shapes for all models of a typical case of each configuration are shown. Finally, some comments on the modeling and analysis of specific realistic systems are made.</p> / Master of Engineering (ME)

<p>The design of column cross sections for known axial loads and moments has reached the stage where practical methods give results which agree very closely with tests and with accurate analyses. However considerable uncertainty exists with regard to methods employed to take into account the effects of the additional moments caused by deflection of columns. Theoretical calculations can be used to accurately predict the loads at which material failure or column instability will occur. However designers require simpler techniques which are sufficiently general in nature to be equally applicable to the large variety of design cases.</p> <p>The effect of column slenderness which is further complicated by consideration of creep under sustained load is the main topic of this study. It is suggested that a realistfc appraisal of design methods must be based on the idea of consistent safety factors. Thus slender columns subjected to sustained load must retain sufficient reserve capacity so that failure loads when compared to design loads provide equal safety factors. The National Building Code of Canada is being revised to include tho relevant provisions of tho ACI Standard 318-71(²). Tho columns analysed in this study were designed in accordance with ACI Standard 318-71. Comprehensive evaluation of the design parameters in the ACI method is given in this thesis along with conclusions and comments. It was observed that the ACI method does not yield consistent safety factors for the different values of the design parameters. The analyses and conclusions of this study are given in details in chapters (6) and (7).</p> / Master of Engineering (ME)

<p>W-shaped steel members are widely used in various structural applications, such as buildings, bridges, and industrial complexes. The recent trend is to produce W -shaped sections using higher and higher strength steel as a replacement for mild-carbon steel, such as 300W steel. However, the width-to-thickness ratios (b/t or w/t) specified in the current CSA steel design standard (1995) for local buckling strength and ductility are based on studies using 300W steel. An investigation was carried out to study the local buckling behavior, with an emphasis placed on flange buckling of compression members, of W-shaped sections made of high strength steel. First, stress-strain characteristics of high strength steel of 350W, 480W and 700Q steel, along with the mild-carbon 300W steel, were determined using standard tensile tests. Based on these experimental material properties, analysis material models were derived for the finite element analysis. Analysis material models used in the finite element modeling were the multi-linear, modified tri-linear, and modified bi-linear models. The next part of the investigation included tests on stub columns of these selected steel grades having flanges at Class I limit based on the current design standard (CSA 1995). Based on the experimental results from these stub column lests, the applicability of b/t Class 1 limits for high strength steel sections were assessed. The experimental results showed that the reserve capacity (f_u/fy) of all the stub columns were close to the sanle, regardless of the steel grade. However, the ductility of the W-shaped sections corresponding to each grade differed substantially. The 300W and 350W steel displayed much more ductility than the 480W and 700Q steel. It was concluded that the bIt limits of a Class I section, which was based on studies on the 300W steel, is also applicable to the 350W steel, but are not transferrable to the higher strength steels. The third part of this study used the analysis material models determined from the tensile tests into a finite element modeling, where a 9-node "assumed strain" shell element was employed. Due to the symmetric behavior of a stub column, a quarter of a stub column was used to simulate the W -shaped section subjected to uniform compression. Meanwhile, an idealized residual stress with parabolic distribution across the web and flange of the W-shaped section was assumed and incorporated in the finite element modeling. Comparing the finite element analysis results with the corresponding experimental results from the stub column tests, an appropriate material model was selected to be used for further finite element analysis. It was found that the results obtained from the tri-linear and multi-linear models were similar to each other, and the modified bi-linear material model best represented the experimental results. Using this bi-linear material model in the finite element analysis, strength and ductility of W-shaped sections with varying flange b/t ratios but constant w/t ratio were determined from the finite element analysis. Results showed that the 350W steel demonstrated characteristics similar to the 300W steel. However, the 700Q steel possessed very little ductility and reserve capacity. The 480W steel was excluded in this part of the study due to the uncertain results from its tensile test. In all of these steel grades, it was found that ductility and reserve capacity decreased as the bIt ratio increased. From both the experimental investigation and the finite element analysis, the b/t limits in the current design standards were found to be applicable to both 300W and 350W steel, but not to the 700Q steel.</p> / Master of Engineering (ME)

<p>No Abstract available</p> / Master of Applied Science (MASc)