Mechatronics Systems Engineering - Theses, Dissertations, and other Required Graduate Degree Essays

Receive updates for this collection

RF Cavity Tuning Based on Reflected Power Measurements

Author: 
Date created: 
2017-03-03
Abstract: 

RF cavities are resonators and the key structures in particle accelerators. An electromagnetic field within the cavities provides the acceleration field. Within linear particle accelerators, particle bunches travel along a beam pipe and through the aligned cavities. A positive electric field along the beam line results in an acceleration kick of the particle bunch. The speed of the traveling particle bunches is synchronized with the RF field withinthe cavities. Hence, tuning of the natural resonance frequency of an RF cavity is essential for accelerator structures to achieve efficient beam acceleration and to reduce power requirements. Operational cavities are typically tuned using phase comparison techniques. Phase measurement is subject to temperature drifts which renders this technique labor and time intensive. In this thesis, we developed two novel resonance frequency tuning schemes solely depending on the reflected power component of a cavity. The base for control scheme development is a mathematical model of the cavity in terms of the steady state signals. The first control scheme was derived through a Lyapunov analysis and incorporates a gradient estimator of the performance function. The second control scheme is based on sliding mode extremum seeking. Both systems are analyzed in terms of the stability; conditions are provided which guarantee stable system behavior up to twice the cavity bandwidth. A simulation study verifies the derived stability conditions. An experimental test bench, including a room temperature quarter wave cavity, was built to test the control schemes under various conditions. Although both control schemes show similar tuning results, a higher tuning accuracy is obtained by the sliding mode based control scheme. Hence, the latter was chosen to be implemented on two resonators, DTL tanks, of TRIUMF’s ISAC I facility. The system was fully commissioned on both DTL tanks and has been in operation since April 2016. Reflected power, forward power, and tuning position are monitored and analyzed. Long term measurements showed the influence of environmental temperature variations. As the influence of environmental temperaturevariations can be neglected for reflected power measurements, the reflected power based tuning system provides a higher tuning accuracy compared to the traditional phase based tuning system. The start-up time and the need for human oversight are reduced significantly.

Document type: 
Thesis
File(s): 
Supervisor(s): 
Mehrdad Moallem
Department: 
Applied Sciences: School of Mechatronic Systems Engineering
Thesis type: 
(Thesis) Ph.D.

Adaptive synchronization of PR controllers in grid-connected inverters

Author: 
Date created: 
2017-05-01
Abstract: 

This thesis focuses on the development of control schemes for single-phase Voltage Source Inverters (VSIs) to ensure that they meet standards for injection of power into the utility grid. For instance, conventional controllers fail to accurately control power flow if there are fluctuations in the grid frequency. Also they fail to provide high quality output currents if the grid voltage is distorted. The aim of this thesis is to address the above problems while adhering to technical standards for interconnected renewable energy systems. For synchronization of the inverter, a closed-loop filter based on the Internal Model Principle (IMP) was developed and its performance was analyzed in response to frequency variations. To this end, we utilized the perturbation-based extermum seeking algorithm to minimize the error and estimate the grid frequency. The designed adaptive controller and filter can estimate grid frequency and achieve a high Power Factor (PF). The effect of harmonic distortion on the control system was investigated and the control scheme was modified to provide low Total Harmonic Distortion (THD) output current. Furthermore, the effect of the DC-link ripple on the PI voltage control loop is analyzed and the control system was modified to attenuate the unwanted third harmonic component in the output current. Simulations were performed using Matlab/Simulink and the digital controller was implemented using Matlab Embedded Coder. A power electronics prototype was built and used to validate the performance of the controller. Based on experimental results, the controller successfully regulates the output power if the grid frequency changes. Also it is able to provide high quality current if the grid is polluted with unwanted harmonic components.

Document type: 
Thesis
File(s): 
Supervisor(s): 
Mehrdad Moallem
Department: 
Applied Sciences: School of Mechatronic Systems Engineering
Thesis type: 
(Thesis) M.A.Sc.

A passive optical proximity sensor

Author: 
Date created: 
2019-11-25
Abstract: 

A new optical proximity sensor was developed and studied in this project. A 3D printed model is used for proving the concept. An integrated model with a new geometry was then microfabricated to further improve the sensor model’s performance. The sensing operation is based on measurement of light intensity falling on photodiodes placed at an inclined angle with respect to the base surface of the sensor. A mathematical model and subsequent experiments prove that it is possible to determine the distance between a light source and the sensor. This sensor is operated passively which means it does not need an active emission for range sensing. The sensor shows reliable operation for short range proximity detection in the range of 5-15 cm. The sensor structure is pyramidal placed flat on the surface with a fixed base angle. Two independent photodiodes are formed on two of the opposite sides of the pyramid. One such pyramid pixel is able to measure the light intensity and the angle at which the light is incident towards the sensor. Using the intensity measured by micro-fabricated test pyramids structures, the distance to the light source is measured. The experimental values demonstrate that the measurements are accurate and repeatable, more so, the device utilizes no active emission to attain proximity measurements. The discussed device can be used for close and continuous proximity detection in mobile devices with low power consumption.

Document type: 
Thesis
File(s): 
Supervisor(s): 
Behraad Bahreyni
Department: 
Applied Sciences: School of Mechatronic Systems Engineering
Thesis type: 
(Thesis) M.A.Sc.

Knowledge assisted metamodeling and optimization method for large-scale engineering design

Author: 
Date created: 
2019-08-22
Abstract: 

Simulation-based design optimization methods commonly treat simulation as a black-box function. An approximation model of the simulation, called metamodel, is often built and used in optimization. However, modeling and searching in an unknown design space lead to high computational cost. To further improve the efficiency of optimization, knowledge of design problems needs to be involved in assisting metamodeling and optimization. This work endeavors to systematically incorporating knowledge for this purpose. After extensive review, two types of knowledge, sensitivity information and causal relations, are employed in solving large scale engineering design problems. Instead of constructing a complete metamodel, a Partial Metamodel-based Optimization (PMO) method is developed to reduce the number of samples for optimizing large-scale problems, using Radial Basis Function-High Dimensional Model Representation (RBF-HDMR) along with a moving cut-center strategy. Sensitivity information is used to selectively model component functions in a partial metamodel. The cut center of a HDMR model moves to the current optimum at each iteration to pursue the optimum. Numerical tests and an airfoil design case show that the PMO method can lead to better optimal results when the samples are scarce. Causal graphs capture relational knowledge among design variables and outcomes. By constructing and performing qualitative analysis on a causal graph, variables without contradiction can be found, whose values can be determined without resorting to optimization. The design problem can thus be divided into two sub-problems based on impact of variables. This dimension reduction and decomposition strategy is applied to a power converter design and an aircraft concept design problem with significantly improved efficiency. Combing the structure of Artificial Neural Networks (ANNs) with causal graphs, a causal-ANN is developed to improve the accuracy of metamodels by involving knowledge. The structure of causal graphs is employed to decompose an ANN into sub-networks. Additionally, leveraging the structure of causal-ANN and theory of Bayesian Networks, the attractive variable subspaces can be identified without additional simulation. Finally, the causal-ANN is applied in a residential energy consumption forecasting problem and both the modeling accuracy and efficiency are improved. This work systematically and methodically models and captures knowledge and brings knowledge in metamodeling and optimization. Sensitivities and causal relations have been incorporated in optimization strategies that have been successfully applied to various engineering design problems. Further research can be extended to studies on how to incorporate other types of knowledge to assist metamodeling and optimization.

Document type: 
Thesis
File(s): 
Supervisor(s): 
Gary Wang
Department: 
Applied Sciences: School of Mechatronic Systems Engineering
Thesis type: 
(Thesis) Ph.D.

Sorption thermal energy storage for sustainable heating and cooling

Author: 
Date created: 
2019-07-12
Abstract: 

Heating and cooling of residential buildings account for 15% of the total energy use in Canada and produce 11% of the total GHG emissions, due to reliance on fossil fuels. Renewable thermal energy and usage of low-grade waste heat offer solutions for decarbonization of heating and cooling. Inherent intermittent nature of such energy resources makes integration of thermal energy storage (TES) systems inevitable. High energy storage density, low heat loss, and using non-toxic and non-polluting refrigerants make sorption TES (S-TES) more appealing and effective for heat/cold storage, compared to other thermal storage methods. This PhD research is set out to assess the performance of low-grade heat-driven S-TES systems for space heating and cooling. As such, the focus of this study is on the thermal and sorption characterization of the sorber bed, mathematical S-TES system modeling, and experimental testing of an S-TES prototype. An analytical model is developed for prediction of thermal conductivity and thermal resistance of packed bed sorbers. Thermal conductivity of packed bed sorber of AQSOA FAM-Z02 with different numbers of layers is measured by heat flow meter for the first time. The model, which is validated by the experimental data, provides a comprehensive platform for the design of packed bed S-TES to (i) predict thermal conductivity and thermal contact resistance of packed bed under the target operating condition and (ii) optimize the packed bed by finding the optimum particle size and arrangement. Small-scale characterizations and screening of sorbent candidates are performed by thermogravimetric analysis/differential scanning calorimetry. Moreover, comprehensive experimental studies are carried out on a custom-built lab-scale S-TES in our lab to study storage performance under various conditions, namely, i) coated vs loose grain sorbent configurations, ii) various heat storage durations, iii) adding high conductive additives in the sorbent material, iv) different operating temperatures, and v) different discharge-to-charge time ratios. A comprehensive transient resistance-capacitance lumped-parameter model is developed to assess the performance of a closed S-TES system. The model is proved to be accurate in comparison with the experimental data and offers a reliable platform for the design and optimization of an S-TES system.

Document type: 
Thesis
File(s): 
Supervisor(s): 
Majid Bahrami
Department: 
Applied Sciences: School of Mechatronic Systems Engineering
Thesis type: 
(Thesis) Ph.D.

The mechanics of contusion spinal cord injury: Towards patient-specific assessments of mechanical loading and injury

Author: 
Date created: 
2019-07-29
Abstract: 

Computational models are becoming an important tool for spinal cord injury (SCI) studies, specifically for transferring in-vivo preclinical achievements to clinical trials and injury prevention design. Despite this, spinal cord tissue properties, constitutive models, and the correlations between tissue mechanics and injury are unclear. Therefore, the anisotropic behaviour of spinal cord tissue was characterized in a human-like animal, constitutive models were employed in SCI computational models, and correlations between SCI model outcomes and tissue injury were evaluated in patient-specific models. Cervical spinal cords were harvested from nine non-human primates (NHPs). White matter samples were cut from lateral columns of the spinal cord. Samples were characterized under dynamic compression. The obtained model was combined with published in-vitro tensile response to capture the anisotropic behaviour of the spinal cord. The model was used to generate subject-specific finite element (FE) models of NHP in-vivo contusion SCI. Capability of several mechanical metrics in predicting tissue damage were evaluated using logistic analysis. NHP spinal cord white matter compressive response was sensitive to strain rate and showed substantial stress relaxation. An Ogden model best captured the white matter behaviour in a quasi-linear viscoelastic material model. Rapid relaxation and high strain rate sensitivity of the white matter indicate that incremental movement of the spinal cord during reparative interventions could substantially reduce the risk of ischemic injury. A fiber-reinforced conditional constitutive model best captured white matter anisotropy. Von-Mises and Tresca stresses showed the strongest correlations with damage in the gray matter. FE tissue damage thresholds were subject-specific except for white matter axonal strain and strain energy density. In summary, the work described herein indicate that measures of mechanical FE outputs correlate with tissue damage both in white and gray matters of spinal cord, and that subject-specific models of human-like animals that include spinal cord anisotropy are able to accurately mimic the biomechanics of contusion impacts.

Document type: 
Thesis
File(s): 
Supervisor(s): 
Carolyn Sparrey
Department: 
Applied Sciences: School of Mechatronic Systems Engineering
Thesis type: 
(Thesis) Ph.D.

Characterization of degradation, fracture and failure in fuel cell membranes

Author: 
Date created: 
2019-06-21
Abstract: 

Proton exchange membrane (PEM) fuel cells offer a promising clean energy solution for the stringent environmental challenges currently faced by the transportation sector. Ionomer membrane is a key component of this technology that enables critical functions such as protonic conduction, electronic insulation, and reactant separation. During operation, however, various chemical, mechanical and thermal stressors degrade the membrane, which compromises the performance and longevity of a fuel cell. A detailed understanding of these degradation processes and their associated membrane failure modes is, therefore, required to develop more durable and economically viable fuel cell systems. In this work, a comprehensive characterization of fuel cell membrane degradation and failure is performed, with a particular emphasis on membrane fracture, through a series of studies encompassing experimental mechanics, accelerated stress testing, microstructural characterization, and numerical modelling under scenarios relevant to the operational environment. Uniaxial tensile fatigue tests are conducted on double edge notch tension (DENT) specimens to measure fatigue crack growth rates in the ionomer membrane and its electrode-coated composite as a function of temperature, humidity and applied stress. The experimental results are further utilized to develop a Paris law based semi-empirical fracture modelling framework to simulate the crack growth rates while numerically implementing the characteristic time-, temperature- and humidity-dependent elastic-viscoplastic constitutive response of these materials. A laboratory-based X-ray computed tomography (XCT) system is utilized to introduce a novel 3D failure analysis methodology for characterizing post-mortem membrane degradation. This methodology is implemented across a series of accelerated stress tests involving both isolated and conjoint chemical and mechanical membrane degradation, respectively, to explain various membrane failure modes and their mechanisms in relation to the nature of in situ stressors, changes in material properties, and influence of neighbouring components. The non-destructive characteristics of XCT imaging are further leveraged to uniquely investigate the 3D structural evolution of a mechanically degrading membrane through periodic visualization of identical locations inside a custom-developed small-scale fuel cell. Electrode cracks and interfacial delamination are identified as critical defects influencing membrane crack initiation, and direct measurements of in situ crack propagation rates are obtained for the first time.

Document type: 
Thesis
File(s): 
Supervisor(s): 
Erik Kjeang
Nimal Rajapakse
Department: 
Applied Sciences: School of Mechatronic Systems Engineering
Thesis type: 
(Thesis) Ph.D.

Sorbent based enthalpy recovery ventilator (SERV) in northern building applications

Author: 
Date created: 
2019-06-03
Abstract: 

Sorbent-based enthalpy recovery ventilator (SERV) is a potential replacement for conventional heat or enthalpy recovery ventilator (HRV/ERV) that require defrosting mechanisms in cold climates, such as in Canada. Sorbent materials (e.g. silica gel, CaCl2, alumina oxide) are non-toxic, inexpensive materials. However, the bulkiness, high pressure drop and large mass of adsorbent are major disadvantages of SERV in packed bed form. In this study, a novel design of sorbent discs with air channels is investigated which feature high heat and mass transfer performance with low pressure drop. A theoretical model is developed for heat and mass transfer in air channels in sorbent discs. A sensitivity analysis performed on design parameters e.g. channel diameter and spacing to achieve an optimum design. A prototype SERV is built in our laboratory. A custom-made experimental set-up equipped with thermocouples, humidity sensors, and an orifice plate air flow meter is designed based on ASHRAE 84 standards to evaluate the performance of the SERV prototype. The performance of the SERV is evaluated for several air flow rate, cycle time and outdoor air temperatures down to -15°C. It is shown that the proof of concept SERV consisting of 2.5 kg of heat storage materials and 2.1 kg of active sorbent material can recover up to 70% of heat and 80% of moisture from exhaust air (up to 20 CFM). It corresponds to 103 W of heat and 43 g of moisture recovery per hour which is comparable to the packed bed sorbent system reported in the literature, however, the proposed SERV offer a 60% less pressure drop.

Document type: 
Thesis
File(s): 
Supervisor(s): 
Majid Bahrami
Department: 
Applied Sciences: School of Mechatronic Systems Engineering
Thesis type: 
(Thesis) M.A.Sc.

Utilizing avalanche breakdown for stress measurements on micro-structures

Date created: 
2019-05-15
Abstract: 

This thesis reports on the usage of the breakdown voltage of a p-n junction diode to measure the mechanical stress/strain in micro-resonators. The working principle relies on the dependence of silicon band gap to the mechanical stress which then affects the current-voltage characteristics of the p-n junction. To explore the effects of mechanical stress/strain on breakdown voltage, a flexural-mode micro-resonator is designed by defining a p-n junction at the anchoring region to experience maximum stress during mechanical excitations. An analytical model has been developed for the study and numerical analysis of this phenomenon. The Synopsys Sentaurus TCAD simulations were employed for the investigation of the breakdown voltage dependence to various mechanical stress magnitudes as well as orientations. A micromechanical device with integrated junctions was designed and fabricated for the validation of postulate. Mechanical stress was applied onto the substrate by subjecting it to mechanical vibrations. It is estimated that the breakdown voltage of the device exhibited a high-stress sensitivity of about 240µV/MPa. The mechanical stress can also be measured by monitoring the device current while biased at a constant voltage. In this mode, the steep changes of the junction current in breakdown region led to nearly tenfold higher stress sensitivity compared to a piezoresistive sensor. The high sensitivity, simple measurement, and potential for miniaturization for breakdown voltage sensing make it a promising technique for measurement of stress in micro- and nano-mechanical devices.

Document type: 
Thesis
File(s): 
Supervisor(s): 
Behraad Bahreyni
Department: 
Applied Sciences: School of Mechatronic Systems Engineering
Thesis type: 
(Thesis) M.A.Sc.

Active power decoupling achieving optimum capacitance requirement with minimal compromise in efficiency

Author: 
Date created: 
2019-04-30
Abstract: 

In single-phase AC/DC converters that achieve unity power factor (UPF) at the AC-side, the power waveform contains a large component at the double-line-frequency (DLF), in addition to the average power. This DLF ripple power can have serious undesirable effects on the load in different applications. In order to prevent it from flowing into the load, the DLF ripple power can be mitigated by connecting a capacitor to the DC-link. However, this method, called passive power decoupling, requires large values of capacitance to be used. For 400(VDC)/kW-level applications, it can be only realized using electrolytic capacitors, resulting in low power-density and low reliability. For applications in which power-density and reliability are more critical, an alternative solution is active power decoupling (APD). In active power decoupling, the storage capacitor has a higher utilization factor because the voltage across it is allowed to have larger variations. This situation can be made possible by separating the capacitor from the DC-link by means of a power electronic converter. The DC/DC buck is the simplest converter that can be used for this purpose. The problem with the buck APD is that it cannot use the theoretically minimum required value of capacitance; although many alternatives have been proposed in the literature, they all bring their own sets of disadvantages. In this work, a superior solution is introduced as an improvement to the buck topology which allows utilization of the theoretically minimum required capacitance with minimal compromise in efficiency. The design details and benefits assessment of this solution are elaborated and its operation is verified using both computer simulation and experiment.

Document type: 
Thesis
File(s): 
Supervisor(s): 
Jiacheng (Jason) Wang
Department: 
Applied Sciences: School of Mechatronic Systems Engineering
Thesis type: 
(Thesis) M.A.Sc.