Thursday 31 January 2013



Date of Award

2010

Degree Type

Thesis

Degree Name

Master of Applied Science (MASc)

Department

Mechanical Engineering

First Advisor

Mohamed Wahab

Abstract

Data Envelopment Analysis (DEA) is a nonparametric optimization technique that evaluates the relative efficiency of decision-making units and is used in this thesis as an empirical estimator of credit rating. The purpose of this research is to combine different DEA models and technique and obtain the best model that captures different aspects of credit risk. Various models are evaluated by combining four Slack DEA models with Principal Component Analysis (PCA), Absolute Weights Restriction, and Stochastic DEA.
We found that Goal Vector Approach Stochastic PCA (SGV+PCA), applied to a sample consisting of five sectors, is the best model. SGV+PCA DEA model obtains a high correlation with Standard & Poor’s (S&P) credit rating and with Market Price; it also classified twelve bankrupted companies within the 17% of the less efficient companies in the sample, suggesting that the model is a good financial health estimator and is a potential tool for credit rating analysis.


Date of Award

2010

Degree Type

Thesis

Degree Name

Master of Applied Science (MASc)

Department

Mechanical Engineering

First Advisor

Greg Kawall

Abstract

An acoustic chamber was designed for testing structure-borne sound transmission in a double-panel assembly induced by point connectors. Several vibration isolators were tested and the overall effects on the noise transmitted through the assembly were predicted by establishing the link between the vibratory acceleration level (VAL) and the sound pressure level (SPL). A detailed assessment of the acoustic chamber showed that a major modification of the double-panel assembly is required before the acoustic performance of this assembly could be evaluated directly using insertion loss (IL) measurements where the sound pressure level (SPL) difference is the performance indicator. This thesis describes the assessment findings and retrofitting options. It is concluded that adjustments to the VAL-to-SPL relation are required to account for distance, radiation efficiency, and room effects. Further adjustments to the acoustic chamber are required to enhance its performance.


Date of Award

2010

Degree Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Mechanical Engineering

First Advisor

Saeed Zolfaghari

Abstract

This dissertation aims to develop an effective and practical method to forecast chaotic time series. Chaotic behaviour has been observed in the areas of marketing, stock markets, supply chain management, foreign exchange rates, weather forecasting and many others. An effective forecasting model can reduce the potential risks and uncertainty and facilitate planning and decision making in chaotic systems. In this study, residual analysis using a combination of the embedding theorem and ensemble artificial neural networks is adopted to forecast chaotic time series. Based on the embedding theorem, the embedding parameters are determined and the time series is reconstructed into proper phase space points. The embedded phase space points are fed into the first neural network and trained. The weights and biases are kept to predict the future values of phase space points and accordingly to obtain future values of chaotic time series. The residual of the predicted time series is further analyzed; and, if a chaotic behaviour is observed, then the residuals are processed as a new chaotic time series and predicted. This iterative residual analysis can be repeated several times depending on the desired accuracy level and computational efficiency. Finally, the last neural network is trained using neural networks' result values of the time series and the residuals as input and the original time series as output. The initial weights and biases of the neural networks are improved using genetic algorithms. Taguchi's design of experiments is adopted to identify appropriate factor-level combinations to improve the result of the proposed forecasting method. A systematic approach is proposed to improve the combination of ensemble artificial neural networks and their parameters. The proposed methodology is applied to a number of benchmark and some real life chaotic time series. In addition, the proposed forecasting method has been applied to financial sector time series, namely, the stock markets and foreign exchange rates. The experimental results confirm that the proposed method can predict the chaotic time series more effectively in terms of error indices when compared with other forecasting methods in the literature.


Date of Award

2010

Degree Type

Thesis

Degree Name

Master of Applied Science (MASc)

Department

Mechanical Engineering

First Advisor

Donatus Oguamanam

Abstract

A finite element model is developed to predict the vibrational response of a single conductor with a Stockbride damper. The mathematical model accounts for the two-way coupling between the conductor and the damper. A two-part numerical analysis using MATLAB is presented to simulate the response of the system. The first part deals with the vibration of the conductor without a damper. The results indicate that longer span conductors without dampers are susceptible to fatigue failure.
In the second part, a damper is attached to the conductor and the effects of the excitation frequency, the damper mass, and the damper location are investigated. This investigation shows that the presence of a properly positioned damper on the conductor significantly reduces fatigue failure.


Date of Award

2010

Degree Type

Thesis

Degree Name

Master of Applied Science (MASc)

Department

Mechanical Engineering

First Advisor

Marcello Papini

Second Advisor

J.K. Spelt

Abstract

Abrasive jet micro-machining is a process that utilizes small abrasive particles entrained in a gas stream to erode material, creating micro-features such as channels and holes. Erosion experiments were carried out on aluminum 6061-T6, Ti-6A1-4V alloy, and 316L stainless steel using 50 μm A1₂O₃ abrasive powder launched at an average speed of 106 m/s. The dependence of erosion rate on impact angle was measured and fitted to a semi-empirical model. The erosion data was used in an analytical model to predict the surface evolution of unmasked channels machined with the abrasive jet at normal and oblique incidence, and masked channels at normal incidence. The predictions of the model were in good agreement with the measured profiles for unmasked channels at normal and oblique impact, and masked channels in at normal incidence up to an aspect ratio (channel depth/width) of 1.25. For the first time, it has been demonstrated that the surface evolution of features machined in metals can be predicted.


Date of Award

2010

Degree Type

Thesis

Degree Name

Master of Applied Science (MASc)

Department

Mechanical Engineering

First Advisor

Ziad Saghir

Abstract

In the present study, a two-dimensional numerical simulation was carried out for binary mixture. The influence of micro gravity vibration or acceleration on board Iinternational Space Station and FOTON-M3, influence of different cavities size as well as the effect of the sign of the Soret coefficient (fluid flow, heat and mass transfer and concentration) in the solvent were investigated in detail. It must be noted that based on previous experiences with this investigation using the same mixture and cavity by Saghir and Parsa [1] the CFD modeling was performed up to 8500s. By this time the quasi-steady state has been reached in most of the mixtures. This thermodiffusion experiment using binary mixture at low pressure condition on the ISS cases shows higher error value for concentration profile at the end of the experiment; compare to FOTON-M3 satellite. However, it should be mentioned that all the errors are significant and more than 25 percent; thus, this kind of investigation should be considered for experiments on all of space vehicles.


Date of Award

2010

Degree Type

Thesis

Degree Name

Master of Applied Science (MASc)

Department

Mechanical Engineering

First Advisor

M. Papini

Second Advisor

J. K. Spelt

Abstract

To aid in the materials selection of gas control valves, the solid particle erosion behaviour of twelve metals was investigated using impinging jets of magnetite particles. The erosion rates were measured for two different particle sizes, two different velocities, and six different impingement angles. Scanning electron micrography and EDX (Energy Dispersive X-ray analysis) mapping was used to investigate the erosion mechanisms and the extent of particle embedding. There was no measurable erosion for the Tungsten Carbide samples, even for very long exposure times. For nickel plated steel, the plating was found to delaminate, resulting in a brittle erosive response. For all other tested materials, the measured erosion rates and scanning electron micrographs indicated a ductile erosion mechanism under all conditions considered. The erosion rates were found to fit a semi-empirical erosion model due to Oka et al. [1] well. The most erosion resistant materials were found to be the Solid tungsten carbide (WC) and Solid Stellite 12 and the least erosion resistant materials were A1018 carbon steel nickel plated and A240 Type 410 stainless steel plate. With all other conditions being equal, a larger erosion rate was measured when utilizing the smaller particles, than when the large particles were used. This counter-intuitive result was demonstrated to be due to a combination of effects, including the formation of thicker hardened layer more embedded particles, and more particle fragmentation when utilizing the larger particles.


Date of Award

2011

Degree Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Mechanical Engineering

First Advisor

Shudong Yu

Abstract

Vibration of simulated CANDU fuel bundles induced by coolant flow is investigated in this thesis through experiments and numerical simulations. Two simulated bundles and a hydraulic loop are built to mimic the situation of the fuel bundles located at the inlet of a fuel channel in a CANDU nuclear reactor. Fuel bundle vibration mechanism is investigated through experiments and numerical simulations. The three-dimensional turbulent flow that passes through the simulated bundles is modeled using the large eddy simulation (LES) and solved with parallel processing. The local cross flows induced by the presence of endplates at the inlet location and bundle interface location are investigated. The fluid forces are obtained as excitations for the fuel bundle vibration analysis. A finite element model of the fuel bundles is developed with the endplates modeled using the 3rd order thick plate theory. The response of the inlet fuel bundle to the fluid excitations is solved in the time and the frequency domain. The added mass and the fluid damping are approximated with the theory on the flow-induced vibration of slender bodies in a parallel flow. Measurements are obtained and used to validate the numerical prediction under various operating flow conditions.


Date of Award

2011

Degree Type

Thesis

Degree Name

Master of Applied Science (MASc)

Department

Mechanical Engineering

First Advisor

Siyuan He

Abstract

In this thesis a novel automated micro assembly mechanism is developed. The assembly mechanism utilizes repulsive-force actuators to flip surface-micromachined 2D structures out-of-plane and assemble them into 3D micro devices. The novel micro assembly mechanism is suitable for wafer-level multi-devices batch assembly without external interference. It can assemble 2D structures not only at the vertical position (perpendicular to the substrate) but also at positions at any angle to the substrate. Two approaches, i.e., graphic method and analytical method, are proposed for designing the micro assembly mechanism. Prototypes are fabricated using the PolyMUMPs surface micromachining technology and tested. The experimental results verify the concept of the novel automated micro assembly mechanism. The strength of the assembled 3D structures in terms of withstanding external acceleration is calculated. The calculated result well matches the experimental result which is about 7g. Using the micro assembly mechanism, 1D and 2D rotation micromirrors are designed for various applications.


Date of Award

2011

Degree Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Mechanical Engineering

First Advisor

Alan Fung

Abstract

The first part of this thesis deals with greenhouse gas (GHG) emissions from fossil fuel-fired power stations. The GHG emission estimation from fossil fuel power generation industry signifies that emissions from this industry can be significantly reduced by fuel switching and adaption of advanced power generation technologies. In the second part of the thesis, steady-state models of some of the advanced fossil fuel power generation technologies are presented. The impacts of various parameters on the solid oxide fuel cell (SOFC) overpotentials and outputs are investigated. The detail analyses of operation of the hybrid SOFC-gas turbine (GT) cycle when fuelled with methane and syngas demonstrate that the efficiencies of the cycles with and without anode exhaust recirculation are close, but the specific power of the former is much higher. The parametric analysis of the performance of the hybrid SOFC-GT cycle indicates that increasing the system operating pressure and SOFC operating temperature and fuel utilization factor improves cycle efficiency, but the effects of the increasing SOFC current density and turbine inlet temperature are not favourable. The analysis of the operation of the system when fuelled with a wide range of fuel types demonstrates that the hybrid SOFC-GT cycle efficiency can be between 59% and 75%, depending on the inlet fuel type. Then, the system performance is investigated when methane as a reference fuel is replaced with various species that can be found in the fuel, i.e., H₂, CO₂, CO, and N₂. The results point out that influence of various species can be significant and different for each case. The experimental and numerical analyses of a biodiesel fuelled micro gas turbine indicate that fuel switching from petrodiesel to biodiesel can influence operational parameters of the system. The modeling results of gas turbine-based power plants signify that relatively simple models can predict plant performance with acceptable accuracy. The unique feature of these models is that they are developed based on similar assumptions and run at similar conditions; therefore, their results can be compared. This work demonstrates that, although utilization of fossil fuels for power generation is inevitable, at least in the short- and mid-term future, it is possible and practical to carry out such utilization more efficiently and in an environmentally friendlier manner.


Date of Award

2011

Degree Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Mechanical Engineering

First Advisor

Liping Fang

Second Advisor

Ling Guan

Abstract

Personalized online systems for Web search, news recommendation, and e-commerce are developed. The process of personalization of online systems consists of three main steps: determining a user's needs, classifying products or services, and matching the user's needs with suitable products or services. A multi-feature based method to automatically classify Web pages into categories of topics hierarchically representing the Web pages is proposed. An approach to modeling and quantifying a user's interests and preferences using the user's Web navigational data is presented. The approach is based on the premise that frequently visiting certain types of content or Web sites indicates that the user is interested in related content or retrieving information from those sites. A personalized search system utilizing a Web user's interest, preference and search context models is developed. A Web user's interest and preference models are constructed and updated by analyzing the user's navigational data and automatically classifying Web pages. A user's search context model is used to determine how the user's interest and preference models impact on his or her search behavior. An algorithm to re-rank search results generated by a conventional search engine is designed to provide a personalized Web search service. A hybrid recommender system of personalized recommendation of news on the Web is developed. Based on the similarities between Web pages and users' models of interest and preference, the Web pages are recommended to the users who are likely interested in the related topics. Moreover, the technique of collaborative filtering is employed, which aims to choose the trusted users and incorporate machine intelligence combined with human efforts. Once trusted users are determined, their behavior on the Web is considered as the manual recommendation part of the system. A method of classifying Web customers for planning customized e-marketing is proposed. The proposed e-marketing approach can be divided into four steps: determining a customer's general interest model, ascertaining a customer's local browsing model, classifying Web customers, and designing a personalized marketing and promotion plan for e-commerce based on the customer classification. Various experiments are carried out to demonstrate the effectiveness of the proposed approaches and systems.


Date of Award

2011

Degree Type

Thesis

Degree Name

Master of Applied Science (MASc)

Department

Mechanical Engineering

First Advisor

C. Ravindran

Abstract

Magnesium (Mg) alloys present a promising alternative to aluminum (Al) alloys in lightweight applications. However, relative to Al alloys, Mg alloys have poor castability. Castability is influenced significantly by the dendrite coherency point (DCP), which represents the temperature, time, and solid fraction at which an interlocking solid network forms during solidification. An increase in the solid fraction at coherency may improve the castability of the alloy and reduce casting defects such as porosity, hot tears and misruns. A successful method for increasing the solid fraction at the DCP in Al alloys involves the use of grain refiners such as titanium (Ti). However, the influence of Ti refiners on the DCP in Mg alloys has not been thoroughly investigated. The objective of this research was to study the effect of Al-5Ti-1B refiner on the dendrite growth mechanism, DCP and porosity of AZ91E magnesium alloy. This thesis is a pioneering effort in relating the grain refinement effect of Ti on the DCP, coherency solid fraction, and porosity development during the solidification of Mg alloy, AZ91E. It represents an important step in improving the castability of Mg alloys. Varying levels of Al-5Ti1B grain refiner (0.005, 0.05, 0.1, 0.2, and 0.3 wt.% Ti) were added to AZ91E. The effect of Al-5Ti-1B grain refiner on the microstructure and dendrite growth mechanism of AZ91E was investigated. Quench experiments were performed to observe transformations in the dendritic morphology that resulted from the refiner additions. The growth rate and DCP were determined using the rheological method. The changes in porosity levels were determined for the grain refiner additions.