This thesis reports theoretical and experimental investigations carried out to better understand the effect of process parameters on the microstructure of transient liquid phase (TLP) joint. The theoretical investigations were carried out using analytical and numerical models to simulate base metal dissolution and isothermal solidification stages of the TLP bonding process. The experimental investigation was carried out by using standard metallographic technique to study the microstructure of bonded materials using optical and Scanning Electron Microscopes. Deviation from parabolic relationship between solid/liquid interface migration and holding time during TLP bonding is suggested as a new alternate phenomenon responsible for the anomalous increase in processing time required to produce eutectic microconstituent free joint with increase in bonding temperature. The results of TLP joining of commercial pure nickel using a Ni-P filler alloy showed that an increase in bonding temperature would be beneficial provided that sufficient holding time is allowed for complete isothermal solidification of liquated insert. Otherwise, an increase in bonding temperature may result in formation of thicker deleterious eutectic along the TLP joint. Furthermore, it was observed that the joint centerline eutectic product and interface second phase particles that form during TLP bonding of Inconel 738 using Ni-P filler can be significantly reduced by post bond heat treatment. The effectiveness of this approach, however, requires proper selection of heat treatment temperature above Ni-P binary eutectic temperature.

This thesis presents a linearized technique to determine a risk-based index for dynamic security. The method is an extension to an existing technique in which the risk of steady state security is calculated using the mean and variance of load uncertainty. The proposed method is applied to calculate the risk indices for the New England 39 bus test system. The results obtained from the proposed method are validated against those estimated by Monte Carlo simulation. Both approaches produce virtually the same results for small load deviations.

With the increased complexity of modern power systems, it may be required more than one platform to do an intended study efficiently and accurately. This research was carried out to investigate the use of co-simulation in an application of power system. A back-to-back Voltage Source Converter (VSC) transmission system was modeled in PSCAD/EMTDC which is an Electro-Magnetic Type (EMT) software. Results were analyzed for some operating points of the system. Then the control system of the above system was modelled in MATLAB/SIMULINK while the rest of the system was modeled in PSCAD/EMTDC. Both of these systems were interfaced to obtain the complete system and results were analyzed under same operating points as the original PSCAD case.

This thesis presents a new way of combining non-linear optimization algorithms and electromagnetic transient (EMT) simulation. In this new combination, a non-linear optimization algorithm utilizes a real-time EMT simulation environment as objective function evaluator. However, as more applications of the off-line EMT simulation software implementation were made, the combination between non-linear optimization algorithms and off-line EMT simulation software revealed new need, which this research work attempts to address. The first need arose from the speed of simulation of the off-line EMT simulation software. With a certain breed of non-linear optimization algorithms, heuristics bases algorithms in particular, a large number of objective function evaluations are required before the termination or convergence criterion in the selected algorithms is satisfied. Sometimes, the number of evaluations as well as the complexity of the simulation case where the objective function is based upon translates into a very long simulation time, which goes beyond the boundary of given resources. This research work attempts to address this simulation timing issue by capitalizing on the real timeness of the simulation environment as well as utilizing the multiple instances of the simulation environment in parallel. The second need arose from the modeling requirement of the off-line EMT simulation software. In order to properly design the necessary objective function evaluator, which is largely a simulation case, a large amount of information needs to be embedded into the case. Under certain circumstances, the necessary information would not be available. Therefore, the simulation case needs to include approximations which may cause compromise in the end result. This limitation becomes more obvious when a real world device such as a commercial controller becomes involved. On the contrary, this limitation can be addressed by the real-time simulation environment because this environment can be directly interfaced with the real world device. In this way, the need for detailed information regarding the device is eliminated. This elimination would enlarge the application of the combination, between the non-linear optimization algorithm and EMT type simulation environment. The effectiveness of the proposed approach is demonstrated by various examples throughout this thesis.

A suite of NLFEA programs, VecTor, has been developed at the University of Toronto. However, this software still requires the development of other functions to execute some types of analyses. One of the required functions is the consideration of transient creep strain in the heat transfer analysis. Moreover, there is a strong need to develop a general graphics-based post-processor applicable to VecTor programs. The first objective of this thesis is to develop a function considering the effect of the transient creep strain, because it can have significant influence on the behaviour of concrete under elevated temperatures. The second purpose of this thesis is to construct the new analysis visualization features compatible with entire suite of VecTor programs. As the result, the modified post-processor, JANUS, has had its abilities expanded significantly.

A suite of NLFEA programs, VecTor, has been developed at the University of Toronto. However, this software still requires the development of other functions to execute some types of analyses. One of the required functions is the consideration of transient creep strain in the heat transfer analysis. Moreover, there is a strong need to develop a general graphics-based post-processor applicable to VecTor programs. The first objective of this thesis is to develop a function considering the effect of the transient creep strain, because it can have significant influence on the behaviour of concrete under elevated temperatures. The second purpose of this thesis is to construct the new analysis visualization features compatible with entire suite of VecTor programs. As the result, the modified post-processor, JANUS, has had its abilities expanded significantly.

Deflocculation is a common upset event in biological wastewater treatment plants and causes significant problems in biosolids discharge and environmental management. However, fundamental understanding of deflocculation is limited. The overall objective of this work was to explore the fundamentals for deflocculation under transients of short-term low dissolved oxygen (DO). The investigation was carried out in a sequence of batch and continuous experiments on activated sludge, followed by batch experiments on E. coli suspensions. Both batch and continuous experiments on activated sludge demonstrated deflocculation of bioflocs under the transients of low DO (< 0.5 mg/L). Under the short-term low DO (in hours), turbidity increased by 20 times in the batch system and by 1-2 times in the continuous system, concentrations of suspended solids increased by 1-2 times, number of small particles (< 12.5 mm) increased by 2 times, more soluble EPS (proteins and humic substances) were released into supernatant or treated effluents, the removal efficiency of organic compounds was reduced by 50-70%. A 40% of increase in bulk K+ but a 30% of decrease in bulk Ca2+ under the DO limitation were observed in the batch experiments. There were significant increases in bulk K+ and decreases in bulk Ca2+ in the continuous experiments. Reversible changes were observed within 24 hours once the DO stress was removed. Floc strength of the remaining bioflocs after deflocculation increased. Deflocculation under the short-term low DO was consistent with an erosion process. The addition of selected chemicals (i.e., Ca2+, tetraethylammonium chloride, glibenclamide, and valinomycin) did not prevent deflocculation under the short-term low DO. It is proposed that a DO stress causes an efflux of cellular K+ but an influx of extracellular Ca2+, resulting in a decreasing ratio of Ca2+/K+ in extracellular solution and thereby causing deflocculation. The E. coli tests supported that increasing bulk K+ under the DO limit was due to the release of cellular K+ and was a stress response to the DO limitation.

This thesis describes the algorithms and techniques developed for harmonic state estimation and transient state estimation, which can be used to identify the location of disturbance sources in an electrical power system. The previous harmonic state estimation algorithm is extended to include the estimation of time-varying harmonics using an adaptive Kalman filter. The proposed method utilises two covariance noise models to overcome the divergence problem in traditional Kalman filters. Moreover, it does not require an optimal covariance noise matrix of the Kalman filter to be used. The common problems faced in harmonic state estimation applications due to the influence of measurement bad data associated with measurements and the lack of measurement points, hence the system being partially observable, are investigated with reference to the Lower South Island of the New Zealand system. The state estimation technique is also extended to transient state estimation. Two formulation methods are outlined and the development of the proposed methodology is presented. Fault scenarios with reference to the Lower South Island of the New Zealand system are simulated to demonstrate the ability of transient state estimation in estimating the voltages and currents of the unmeasured locations, and applying the estimated results to search for the fault location. The estimation results are compared with PSCAD/EMTDC simulations to justify their accuracy.

This thesis presents the application of graphics processing unit (GPU) based parallel computing technique to speed up electromagnetic transients (EMT) simulation of large power systems. GPUs support extra computing capability to handle gaming and animation related applications in the desktop computers. GPUs can be used for general-purpose computations, such as EMT simulation. Traditionally, EMT simulation tools are implemented on the CPUs, where simulation is performed in a sequential manner. Hence, with the increase in network size, there is a drastic increase in simulation times. This research shows that the use of GPU computing considerably reduces the total simulation time. This thesis proposes parallelized algorithm for EMT simulations on the GPU, and demonstrates the algorithm by simulating large power systems. Total computation times for GPU computing, using 'compute unified device architecture' (CUDA)-based C programming are compared with the total computation times for the sequential implementations on the CPU using ANSI-C programming for systems of various sizes and types. Special parallel processing techniques are implemented to model various power system components such as transmission lines, generators, etc. An advanced technique to implement parallel matrix-vector multiplication on the GPU is implemented, which shows a significant performance gain in the simulation. A sparsity-based technique for the inverse admittance matrix is implemented in this simulation process to ignore the multiplications involving zeros. A typical power electronic subsystem is also implemented in this simulation process, which had not been implemented in the literature so far for the GPU platforms. GPU computing-based simulation of large power networks with many power electronic subsystems has shown a massive performance gain compared to conventional sequential simulations with and without the sparsity technique. Finally, in this research work, the effect of granularity on the speedup of simulation was investigated. Granularity is defined as the ratio of the number of transmission lines used to interconnect various subsystems to the total size of the network. It should be noted that dividing a network into smaller subsystems requires additional transmission lines. Simulation results show that there is a negative impact on the overall performance gain of simulation with the use of excessive transmission lines in the test systems.

No description available.