Search

High energy particles in the outer space could flip the state of the latches of the electronic devices. The upset of the latches is called soft error since it would not cause permanent damage to the device. However, the soft error may cause faults in a processor and lead to malfunction of the computer system on the satellite. In this thesis, we implement a novel soft error fault tolerant scheme based on the LEON2/3 processor to protect the processor from soft errors. We verify the correctness and evaluate the overhead of this scheme, and we also determine the critical resource which should be protected. Our scheme includes two parts: sensor network and rollback scheme. A sensor is used to monitor a target register. It will assert if the monitored register is flipped because of a soft error. The rollback scheme is modified from a synchronization feature of the LEON2 and LEON3 processors. This feature originally aims to synchronize among Floating Point Unit (FPU), CACHE and Integer Unit (IU). We make use of it to stall the IU when a soft error is detected, and recover from the error by re-executing the current operation. To verify the correctness of the sensor and rollback scheme, we inject errors during the execution of a large number of instructions of LEON2/3. The results show that all instruction rollbacks are correct. To evaluate the overhead of the scheme, we determine the time and resource penalties of our scheme. The test results show that the scheme incurs only one extra clock penalty in about 90\% of test cases and increases 0.282% of resource usage of the original processor for adding one 32-bit sensor. To identify the critical resource for protection, we define the weight of instructions based on the frequency of instruction usage. Moreover, we monitor the number of accesses of the internal registers/bits in LEON3. Then we compute the impact factor (IF) for each internal register and status bit according to the register access frequencies and instruction weight. Using this approach, we could figure out the most critical resources according to the impact factor. The results show that there are 233 registers and status bits in total. Of which, 91 of them have 100% IF which suggests they are the critical resources. 43 of them have an IF ranging between 1.9% and 97.6%, which means they are less important. The IF of the remaining 89 is 0 because they are not used in our target application. Our work shows that the sensor network and rollback scheme can protect the processor from soft errors while incurring minimal penalties. By analyzing the vulnerability of registers and status bits, we can selectively deploy limited resource on the most critical registers/bits. Therefore, our scheme could effectively improve the robustness of computer systems for satellite applications. MASTER OF ENGINEERING (SCE)

152 p. The aim of this project is to investigate the suitability of rare-earth oxide as gate dielectric, in particular LU2O3. There are three main chapters, in which the properties of LU2O3 film, the effect of post-deposition treatments on LU2O3 film as well as the properties of Lu203 on the next generation of semiconductor substrate are discussed in detail. DOCTOR OF PHILOSOPHY (MSE)

With the prevalence of coupling engineering systems such as multi-agent systems, wireless networks and the smart grid, solving optimization problems across multiple decision-makers emerges to be of theoretical significance and practical relevance. Resource allocation for systems with limited resources, energy consumption control for electricity users, and economic dispatch among a network of generators in smart grids are typical examples of multi-decision-maker optimization problems. These optimization problems are explored from the perspectives of non-cooperative games and networked optimization in this dissertation. The background, introduction and motivation for this topic are included in the first two chapters. The specific problems considered in the dissertation are summarized as follows. Energy consumption control among a network of electricity users in smart grids is considered. An aggregate game approach is proposed to solve it by noticing that the aggregate energy consumption is unknown to the electricity users during the energy consumption scheduling process. Average consensus based updating strategies are developed to search for the Nash equilibrium of the energy consumption game. A general energy consumption game with possibly multiple Nash equilibria is firstly explored. In this case, the Nash equilibrium is exponentially stable under the given strategy. In particular, energy consumption game (for heating ventilation and air conditioning systems) that admits a unique Nash equilibrium is then investigated. Based on the uniqueness of the Nash equilibrium, a non-local convergence result is derived. Furthermore, energy consumption games with stubborn players are studied. With the existence of stubborn players, the rational players' actions converge to a neighborhood of the best response strategies. By utilizing the proposed methods, the electricity users only need to communicate with their neighbors about their estimates on the averaged aggregate energy consumption. No private information is exchanged among the electricity users. Hence, the proposed methods are free of privacy concern. Since the Nash equilibrium may not be efficient from the system-level perspective, distributed optimization problems are investigated. In Chapter 4, a distributed extremum seeking framework is proposed to solve networked optimization problems. Under some conditions, the proposed method enables the agents' strategies to converge to the optimal solution without using the explicit expressions of the constraints functions, the cost functions or their gradients. Noticing that many engineering systems exhibit time-varying characteristic while continuous-time time-varying distributed optimization remains as an open problem, we consider it for systems with quadratic objective functions in this dissertation. Time-varying distributed optimization with neighboring coupled objective functions is firstly considered. A robust gradient based method is proposed to solve it. More generally coupled distributed optimization problems are then investigated by penalty function based methods. The penalty function based methods can be utilized to approximate the solution of the distributed optimization problem. The last problem considered is seeking the solution to non-cooperative games without using explicit model information. Different from existing works that regard the Nash equilibrium as a ``fixed point", Nash equilibrium is treated as a time-varying trajectory in this dissertation. The time-varying Nash equilibrium seeking problem is addressed by using an extremum seeking approach. Sinusoidal excitation signals are utilized in the extremum seeking loop to modulate and demodulate the players' actions. A delay-based subsystem is employed to estimate the gradient and a robust tracking strategy is designed to search for the Nash equilibrium based on the estimated gradients. For symmetric quadratic games, it is shown that the players' actions converge asymptotically to the Nash equilibrium trajectory. For more general quadratic games, a bounded convergence result is presented. In summary, this dissertation studies optimization problems across multiple decision-makers. If the decision-makers are self-interested, the problem may be modeled as non-cooperative games. Nash equilibrium seeking for non-cooperative games are considered. To achieve the system-level objective, networked optimization is investigated. An extremum seeking approach is proposed to solve it which is followed by some explorations on time-varying distributed optimization problems. To demonstrate the practical relevance of the non-cooperative games and the networked optimization problems, energy consumption control for electricity users in smart grids is studied. DOCTOR OF PHILOSOPHY (EEE)

In Singapore, incineration is the mam waste management technology used for handling municipal solid waste (MSW), in conjunction with recycling and other waste minimization policy. After incineration, the volume of waste can be reduced up to 90%, thereby significantly reducing the demand of land space for the final disposal of waste. However, incineration cannot completely eradicate the need for landfill as the end product of incineration, i.e. incineration ash, still has to be disposed of properly in a sanitary landfill. Currently, Singapore's only offshore landfill, Pulau Semakau, is estimated to last till 2035. With a lack of suitable alternative landfill in land scarce Singapore, there is an urgency to utilize incineration bottom ash (IBA), which accounts for 85-95% of the total residues produced after incineration, so as to prolong the lifespan of the landfill. Before IBA can be utilized, it has to be treated to reduce the leaching of heavy metals. This research proposes accelerated carbonation for IBA treatment as it has some advantages over other methods, and has not yet been explored in Singapore. Literature review has also shown this process is sustainable and economically viable.Before accelerated carbonation was carried out, the physical and chemical characteristics of untreated IBA were studied by collecting and analyzing IBA from two incineration plants in Singapore, over a 6 months period. The two incineration plants, i.e. SWTEP and TSIP, were selected based on the types of waste received, which could then produce different mineralogical compositions of IBA. The 6 months sampling period was decided to cover the possible temporal variation of IBA characteristics. This characterization study showed that the total element contents of IBA from both incineration plants in Singapore are similar to those in other countries reported in literature. However, elements like Cd, Co, Hg, Pb and Se exhibited high variation over the investigated sampling period. Pb showed similar high variation in leaching test. The accelerated carbonation experiment was divided into two parts. In the first part, the influence of moisture content and temperature on accelerated carbonation was investigated using the 0-2 mm size fraction of IBA. The optimum operating conditions found in the first part was then applied to other size fractions (i.e. 2-4 mm, 4-20 mm and 20-50 mm) in the second part of the experiment. Results showed that the optimum operating conditions fixed at 20% CO2 for the two incineration plants were slightly different. For SWTEP, it was 35°C and 15% moisture content. For TSIP, it was 50°C and 15% moisture content. The carbonation duration for both plants was 2 hours. XRD analysis showed slight difference between the mineralogical compositions of the two incineration plants, which could explain the different optimum carbonation temperatures. Most of the heavy metals (i.e. Pb, Zn, Cu and Cr) were found to reduce in leaching after accelerated carbonation, whereas Mo and Sb leaching increased after carbonation. After sieving, the amount released from untreated IBA decreased as the size fractions increased. The leaching mechanism among the different size fractions was found to be similar. Despite the limitations in reducing the leaching of certain trace elements, accelerated carbonation was able to significantly reduce the amount released of Pb, Zn, Cu and Cr, which were much higher in the amount released, compared to Mo and Sb. On the other hand, accelerated carbonation was found to be ineffective in reducing the leaching of soluble salts. In conclusion, this research has shown that accelerated carbonation has the potential to treat IBA effectively within a short time frame. Doctor of Philosophy (CEE)

Fuzzy neural networks is a popular combination in soft computing that unites the human-like reasoning style of fuzzy systems with the connectionist structure and learning ability of neural networks. There are two types of fuzzy neural networks, namely the Mamdani model, which is focused on interpretability, and the Takagi-Sugeno-Kang (TSK) model, which is focused on accuracy. The main advantage of the TSK-model over the Mamdani-model is its ability to achieve superior system modeling accuracy. TSK fuzzy neural networks are widely preferred over their Mamdani counterparts in dynamic and complex real-life problems that require high precision. This Thesis is mainly focused on addressing the existing problems of TSK fuzzy neural networks. Existing TSK models proposed in the literature can be broadly classified into three classes. Class I TSK models are essentially fuzzy systems that are unable to learn in an incremental manner. Class II TSK networks, on the other hand, are able to learn in incremental manner, but are generally constrained to time-invariant environments. In practice, many real-life problems are time-variant, in which the characteristics of the underlying data-generating processes might change over time. Class III TSK networks are referred to as evolving fuzzy systems. They adopt incremental learning approaches and attempt to solve time-variant problems. However, many evolving systems still encounter three critical issues; namely: 1) Their fuzzy rule base can only grow, 2) They do not consider the interpretability of the knowledge bases and 3) They cannot give accurate solutions when solving complex time-variant data sets that exhibit drift and shift behaviors. In this Thesis, a generic self-evolving Takagi–Sugeno–Kang fuzzy framework (GSETSK) is proposed to overcome the above-listed deficiencies of existing TSK networks with the following contributions: A novel fuzzy clustering algorithm known as Multidimensional-Scaling Growing Clustering (MSGC) is proposed to empower GSETSK with the incremental learning ability. MSGC also employs a novel merging approach to ensure a compact and interpretable knowledge base in the GSETSK framework. MSGC is inspired by human cognitive process models and it can work in fast-changing time-variant environments. To keep an up-to-date fuzzy rule base when dealing with time-variant problems, a novel ‘gradual’-forgetting-based rule pruning approach is proposed to unlearn outdated data by deleting obsolete rules. It adopts the Hebbian learning mechanism behind the long-term potentiation phenomenon in the brain. It can detect the drift and shift behaviors in time-variant problems and give accurate solutions for such problems. DOCTOR OF PHILOSOPHY (SCE)

The research work focuses on the design, fabrication, and characterization of novel three-dimensional carbon-based hybrid materials in order to utilize them in next-generation power tools. Through novel functionalization and structuration, several three-dimensional lattices based on graphene or carbon nanotubes are designed and fabricated. A variety of techniques of both materials and chemical science are used to fully characterize the functional hybrids, and evaluate their application potential as anode or cathode for lithium-ion battery. DOCTOR OF PHILOSOPHY (SPMS)

According to the World Health Organization, chronic diseases are estimated to account for 63% of deaths globally. However, adherence to chronic diseases therapy only averages 50% in developed countries. Despite decades of research, nonadherence to medication remains an unresolved healthcare issue. It is often thought that cognitive deficiency is the main reason for medication nonadherence; however, research showed that patients’ active decision not to comply made up around half of the reasons. In the recent years, medication-taking behaviors have been considered from a reasoned decision-making perspective, where the medication nonadherence behaviors are categorized into unintentional or intentional factors. Therefore, a proper framework is necessary to explain intentional and unintentional nonadherence. In this research, a hypothetical model is investigated to explain Intentional and Unintentional nonadherence and validated with subsequent studies. Two studies to understand medication nonadherence attitude in Singapore are carried out. The investigation revealed that only about 25%-30% of respondents are considered highly adherent to their medication in Singapore, based on the Morisky Medication Adherence Scale. About 40%-55% of the respondents cited “Forgetting” (i.e. unintentional nonadherence) for not taking their medication, while intentional nonadherence made up about 20-25% of the reasons. Differences between Younger and Older adults attitude towards medication adherence are also found. Younger adults are found to be more susceptible to both intentional and unintentional nonadherence, and the latter is influenced more by “busyness” than age. We propose an Ability-Belief-Context (ABC) framework for understanding predictors of Unintentional and Intentional Medication Nonadherence. Ten variables and constructs are analyzed to identify relationships with intentional and unintentional nonadherence, through mediation analysis. The variables and constructs are modeled into Ability, Beliefs and Context to study their influence on intentional and unintentional nonadherence: Ability includes: Perceived Prospective Memory, Perceived Ability (to self-manage medication) and Lifestyle Busyness; Belief includes: Trust in Medication, Trust in Doctor and Knowledge; while the Context includes: Social Support and External Cues (e.g. pillbox use). Attitude, Beliefs, Trust in Medication and Trust in Doctor are consistently found to influence intentional medication nonadherence. Perceived Prospective Memory and Busyness are significantly associated with both adherence and attitude, and mediated by both intentional and unintentional factors. Contextual factors, such as Social Support in reminding an individual to take their medication have an effect on unintentional nonadherence. Pillbox use is however not found to contribute to intentional nor unintentional nonadherence. This research made several contributions in the understanding of medication nonadherence: First, a comprehensive study on reasons for medication nonadherence in Singapore is studied and validated through subsequently studies. Second, ten factors contributing to intentional and unintentional are analyzed to determine both intentional and unintentional mediating effects. Third, age-related differences and attitudes towards medication nonadherence are analyzed. Fourth, a combination of psychosocial and human factors approaches is adopted to understand medication nonadherence. And last but not least, a framework is proposed to identify the relationships between predictors, mediating factors of intentional and unintentional nonadherence based on Abilities, Beliefs, Context and Age. Doctor of Philosophy (MAE)

From Google to Pinterest, multimedia search engines such as Google Goggles deliver a wealth of visual information related to the search query. It recognizes and provides useful information when pointing the mobile phone camera at a business card, a book, a painting, a famous landmark, or a barcode. Vision-based techniques try to perceive and understand images by learning from the ability of human vision. Developing such techniques remains an ongoing challenge for computers. Nowadays, multimedia systems for online advertising and commerce have a large market demand. Recent years' computer vision and multimedia communities have devoted efforts on many applications, such as fashion retrieval or recommendation for clothing, shoes, etc. Handbag has become a desirable fashion accessory, with six in ten consumers having purchased at least one new handbag in the year of 2014. Such market demand motivates the handbag recognition related vision products. However, this kind of product is still limited so far. As Google says, Goggles does not work well yet on things like food, plants, animals and some fashion items such as handbags. To develop such reliable recognition engines, we study handbag recognition and recommendation, which are key steps for building up a multimedia search system. The works in this thesis can be summarized as below. A style-to-color discriminative representation framework for handbag recognition is carried out at first. We identify the handbag model by conducting the style-based recognition and color-based recognition sequentially due to the visual characteristics of handbags. Experiments are conducted on our newly constructed handbag datasets. The experimental results illustrate that our method achieves over 10% improvement in accuracy for recognizing handbags when compared with existing fine-grained or generic object recognition methods. In recent years, Convolutional Neural Network (CNN) is promising for many image recognition tasks, which motivates us to design a handbag recognition algorithm based on CNN. However, after studying various CNN architectures for training the classifier, we find that the previous CNN models do not provide discriminative color information during training. Moreover, CNN models usually consider the hard label (i.e., the ground truth class label) to train a multiclass classifier. This is not sufficient especially for visually similar classes. In order to train a better CNN for classification, we present a Feature Selective joint Classification-Regression CNN (FSCR-CNN) model. It is helpful for recognizing color sensitive objects and it facilitates the classifier modeling for visually similar classes. Moreover, we propose an end-to-end handbag recognition framework. In this framework, we propose three components: (1) symmetry-based proposal localization, (2) CNN detection and FSCR-CNN classification, and (3) combination of detection scores and classification scores by conditional probability model. The experimental results verify the advantages of each component of our framework for handbag recognition. A handbag recommendation system for e-commerce and shops is also proposed. It can help shoppers to find desirable fashion items, which facilitates online interaction and product promotion. Given the images of the shopper's preferred handbags, the recommendation is performed by joint learning of attribute projection and one-class SVM classification. A weighted AutoEncoder method is further proposed to refine the recommended results. The results show that this scheme performs favorably based on the initial subject testing. ELECTRICAL and ELECTRONIC ENGINEERING

In recent years, the concept of green building enjoys a great popularity throughout the world. Energy conservation and Indoor Environmental Quality (IEQ) improvement in buildings receive consistent attentions from all walks of life. As the core element to create comfortable and healthy indoor environmental conditions for human beings, Heating, Ventilation, and Air-Conditioning (HVAC) systems which are also the largest source of energy consumption are necessary to be well studied and developed. Among various HVAC schemes, using active chilled beam terminal units is a very superior option for next generation HVAC solutions, which was listed as one of the 15 most promising HVAC related technologies by American Council for Energy Efficient Economy (ACEEE) in 2009. Active chilled beam terminal units based HVAC systems originate in Scandinavia and have been adopted widely in Europe and to some extent in Australia. More recently, the systems are penetrating into North America and Asia. However, in depth investigations on the systems are still inadequate. Some technical difficulties have emerged in the existing engineering application and need to be resolved for a wider acceptance, especially in the emerging markets. Therefore, this thesis tries to put some effort on this front, with the focus on the design optimization, modeling, and performance evaluation of active chilled beam terminal units for the tropical climate. The contributions of the thesis are briefly summarized as below: 1. As a core part of active chilled beam terminal units, the secondary heat exchanger should not be a standard “off the shelf” product as it used to be. In order to ma ximize t he coo ling capac it y w hile min im ize t he e ne r gy cons umpt io n, t he c irc uit a rra nge me nt is opt im ized. A n e xpe r ime nta l comparison study of four fin-tube heat exchangers with different circuit numbers is conducted to determine the optimal circuit number. Then, tube connecting sequences of the circuits are investigated. Though a series of experiment-aided simulations taking the in-situ secondary air velocity profile into consideration, optimal tube connecting sequences are proposed. With the optimized circuit arrangement, the performance of the secondary heat exchanger, as well as the terminal unit is substantially enhanced. More importantly, the findings and used methods will have significant effects on the design of future active chilled beam terminal units. 2. An appropriate model of active chilled beam terminal units is indispensable in the system design, simulation, performance evaluation, as well as development of advanced control and optimization strategies. However, the issue has been so far overlooked by the research communities. In this work, a hybrid dynamic model of the terminal units with few unknown parameters is established by deriving the model using first principles and estimating the parameters experimentally. Through this approach, a reasonable compromise is made between capturing the exact underlying physics and suitability for engineering applications. Static and dynamic performances of the model are verified. As the first reported model of active chilled beam terminal units, it is expected to have a wide range of applications in the aforementioned aspects. In addition, the modeling technique can be extended to the other terminal units. 3. When promoting the application of active chilled beam terminal units in different climates, inappropriate understanding of operating characteristics and efficiencies of the terminal units has probably been the essential obstacle. For example, in tropical regions, not only the sensible cooling capacity but also the latent cooling capacity of the terminal units should be matched with the counterparts of conditioned spaces to strictly avoid condensation. Nevertheless, the latent cooling capacity has never been involved in any studies. In order to address this issue, a series of simulations are carried out. The operating characteristics and efficiencies of the terminal units under variable air volume mode are revealed for the first time. The obtained result will be fundamental in designing and operating active chilled beam systems. DOCTOR OF PHILOSOPHY (EEE)

Steel is widely used in modern building constructions and consequently, extensive research on steel structure is carried out all around the world. Among steel structures, the use of high strength steel is gaining popularity and wider usage. The exact effect of using the high strength steel on the strength of joints has not been investigated extensively so far and there are no published design equations available at present. This study will investigate the influence of steel grade on the strength of Square Hollow Section K-joints. The results from this study mainly will complement the on-going research on the MND-SUL project. In the MND-SUL project, a functional truss bridge across an express way is planned to be built and the mega-truss will be made of S690 steel sections. This thesis presents the numerical studies for assessing the strength of high strength steel SHS gap K-joints. The aim of the numerical analysis is to improve the current design equation in CIDECT Design Guide 3. The results demonstrate the influence of high strength steel on the design resistance of K-joints and two new formulas are proposed to improve the existing equation. Two finite element models are validated with experimental results as the basic verification step. The parametric study is conducted which comprises 60 models of SHS K-joints made of S690, 60 models of gap SHS K-joints made of S355 and 60models of SHS K-joints made of S460. Each model is subjected to 3 load cases. The geometric parameters for the joints such as width ratio between braces and chord (β), half width to thickness ratio of chord (γ), the thickness ratio between braces and chord (τ) are varied in the parametric study. The higher the preload is, the lesser the strength of gap SHS K-joints observed. It is found that chord face failure by yielding or buckling is the most common type of failure for β of 0.6 to 0.8, 2γ of 30 to 35 and τ of 0.79 to 1. The numerical results are compared to existing formula in CIDECT Design Guide 3. It is found that the numerical ultimate capacity of the joints is less than the theoretical results predicted by CIDECT Design Guide 3. Two parametric formulas for predicting the design resistance of gap SHS K-joints with high strength steel are proposed in the thesis. The results of proposed formulae are compared with the numerical results and they show good agreement. MASTER OF ENGINEERING (CEE)

The demand for better efficiency, challenging dynamics, flawless systems and autotuning features has directed the attention of researchers to adaptive controllers. These controllers have been widely developed and implemented for a vast range of applications. If they are well designed, they can improve the performance and simultaneously expand the operational range, in contrast to the trade-off that is commonly observed in robust controllers. Another important feature of these controllers is their capability to compensate for unmodelled dynamics. Furthermore, conventional controllers lack the ability to properly handle changes in the model that could be caused by a failure or environmental variations. The control of unmanned aerial vehicles (UAVs) is one of the many applications in which these controllers are implied and tested. The use of multirotors for civil and military applications is continuously increasing but their unstable dynamics in the absence of a proper controller, in addition to the higher complexity and more challenging dynamics, promote the need for better and flexible control techniques, such as adaptive controllers. The progressive need for the improvement of the quadcopter's performance, minimisation of modelling data and the possibility to integrate fail-safe and autotuning features were the motivation for this research. Although independent command-following and disturbance rejection of constant and harmonic signals were successfully achieved, multiple limitations regarding the adaptability of the controller are found. This thesis provides a description of the design and implementation of a position controller for a quadcopter using a retrospective cost adaptive controller (RCAC). First, the RCAC is briefly introduced and a description of its working principle and main components needed for its computation is given. Secondly, the implementation and validation of the algorithms are analysed and the results are linked with the equations to provide insight into the expected behaviour of the controller. Then, the dynamic model and the structure of the attitude control are described. Finally, the integration of the RCAC for position control is explained and based on the obtained results, an evaluation of the performance and the influence of the parameters is provided. Master of Science (Aerospace Engineering)

There is a surge in demand for biomass/biofuel boilers in Sweden. Iron based coatings and nickel based coatings are proven to be a better at protecting the boilers in the corrosive and harsh environment. The objective of this study is to see how the new High Velocity Air Fuel (HVAF) thermal spray process compares to the traditional High Velocity Oxygen Fuel (HVOF) thermal spray process in the domain of corrosion-erosion protection in a harsh environment and how do iron and nickel based coatings compare to each other. The corrosion behavior was studied when the samples were kept in furnace at 600˚C for different time periods of 12, 48, 96 and 168hrs. One set of samples were kept in ambient air and the other set were coated with KCl solution corresponding to 1.35 μmol K+/cm2. The resultant weight gain is studied and noted, and the oxide layer formed on the coating is studied and the results are concluded. The erosion testing was done with the help of an air jet erosion test rig at room temperature. The net loss in weight due to erosion was calculated and then noted for analysis. It was observed that the HVAF iron based coatings perform much better in term of corrosion and erosion resistance as the weight gain is minimum as compared to the HVOF iron based coating.The weight loss due to erosion is also lower in HVAF iron based coatings. When the iron based coatings were compared with the Nickel based coatings it showed that the Nickel based coatings which were sprayed with HVAF process performed much better in terms of corrosion-erosion performance when compared to HVAF iron based coatings. Thus, HVAF coatings are better than HVOF coatings in terms of corrosion-erosion performance and Nickel based coatings are better than iron based coatings in terms of corrosion-erosion performance. Master of Science (Aerospace Engineering)

The Laser Based Additive Manufacturing is an Additive Manufacturing (AM) process in which a part is built or constructed in layers by depositing materials as per the 3D design model. The monitoring of this process is done to ensure whether the part is built as per model dimensions. However, the monitoring process currently experiences problems due to perspective defects, lack of requirement specific calibration and positioning of the Charged Coupled Device (CCD) camera used. Usually, the correction of perspective is done by using devices which require additional setup for their functioning. Deploying additional instruments require space which creates more constraints in the monitoring system. The algorithmic approach for perspective correction requires great understanding of the concept of planes and 3D geometry. The problem of lack of requirement specific calibration is addressed by algorithms and experimental techniques that does not provide a proper physical calibration pattern for measuring the dimensions of the part and to confirm the validity of perspective correction algorithm. This dissertation involves the reduction or possible elimination of the mentioned problems. Image processing tools such as MATLAB and Adobe Photoshop CS4 are used to propose elegant solutions to overcome the errors caused by perspective issues. A grid paper is introduced as a simple and an effective means of measuring the part's dimensions and verifying the correctness of the perspective correction algorithm. The problem of positioning the CCD is solved by means of conducting an experiment which is based on the concepts of concentric circles. Additionally, this paper discusses the repeatability error of the experiment performed to solve the positioning issue of the CCD camera. Solutions are supported with the appropriate results and inferences. Master of Science (Systems & Project Management)

With the immense popularity and convenience poised by air travel, passengers all over the world have been presented with the opportunity to travel to any parts of the world at any time. lATA recently published a press release forecasting that 7.2 billion people will travel by air by 2035 compared to 3.5 million at present. This calls for the need for tourist destinations like Sri Lanka to be ready in terms of air passenger volume and anticipating the level of infrastructural development required to cater to the increasing demand. The aim of this thesis is to estimate and forecast the air passenger demand that Sri Lanka may entail in the next 15 years. The focus is on finding a suitable forecasting model that presents optimal forecasting accuracy to predict the future. I 0 different forecasting models were experimented and evaluated in the thesis based on various evaluation criteria. The best perfom1ing forecasting model was then selected and the passenger demand was estimated for the next 15 years. Master of Science (Smart Product Design)

Industries such as aerospace, automobile, marine, medical, optical require compliant tools for material removal and for achieving a good surface finish in the intricate shape of the components. Hard tool operations such as milling, turning and conventional grinding are inefficient in machining intricate shapes and geometry. Abrasive belt grinding process is one of the most widely used two-body abrasive compliant machining processes wherein the abrasive belt is forced against the semifinished components of complex shape and geometry to machine off topographies such as burr, chip and to achieve the required material removal and surface finish. An abrasive compliant belt grinding system consists of an abrasive belt with fine grains, a driver unit, a compliant polymer backing of suitable hardness and a robot manipulator is used for tool path planning. Compliant belt grinding tool adapts to a part profile, and a controlled force is exerted on target component to remove material. Belt grinding process is highly nonlinear due to the complexity of the underlying unknown physical mechanism. The presence of multiple parameters working in different regimes creates a dynamic condition which is not entirely understood well in industries. These dynamics arise from the multiple parameters such as belt speed, feed rate, and grit size involved. Most industries operational in belt grinding process are still largely based on empirical rules and experience of the operator. Quantification of such a dynamic process of material removal is not yet studied in detail even in the scientific research works. This research study is conducted to understand the material removal of the aluminium workpiece with different cutting parameters such as belt speed, feed rate, roller rubber property, force, and the grit size of the cutting tool. The inter-relation and prominence of these parameters are studied by performing experimental trials using Taguchi Design of Experiments (DoE). Taguchi design of experiments is adopted here due to its timesaving factors . The trial result is then analysed using Analysis of Variance (ANOVA) to understand the effect of changes in the process parameters to that of material removal. The regression model based on multilayer artificial neural network was developed for belt grinding process to model the material removal process computationally. This study is an attempt to bridge the gap between understanding material removal of belt grinding process and contribution of process parameter involved in abrasive belt grinding process. Master of Science (Precision Engineering)

Recent advances in the field of unmanned aerial vehicles (UAVs) make rotorcrafts equipped with a camera suitable for performing advanced missions. One of the primary applications is tracking stationary and moving targets in GPSdenied environment or when the relative positioning is needed. Visual servoing provides such an alternative. However, image-based visual servoing (IBVS) for a rotorcraft UAV is challenging due to the nonlinear perspective projection of image formation, system uncertainties and underactuated nonlinear dynamics of the UAV. An IBVS law in consideration of system dynamics is known as dynamic image-based visual servoing (DIBVS). Among various DIBVS approaches, this research selects the virtual camera approach which is drawing more attention recently because it facilitates the estimation of image points’ depth and image kinematics. This thesis presents a dynamic output feedback image-based visual servoing law for rotorcraft UAVs. The proposed control law enables a UAV to regulate its position and heading relative to a visual target consisting of multiple coplanar points. The UAV is required to be equipped with a minimal set of sensors, i.e., an inertial measurement unit (IMU) and a single downward facing camera. A set of first order image moment features which are defined in virtual camera image plane are used for visual servoing. The proposed control law has the following characteristics. First, traditional IBVS approaches control the lateral and height subsystems separately and usually do not consider the potential loss of the image. Thus, in this research, the desired value of the image feature in controlling the vertical motion of the vehicle is designed to be a function of the image features of the lateral motion instead of a constant. This modification helps to keep the visual target stay in the camera’s field of view (FoV) by indirectly adjusting the height of the vehicle. Second, the dimension of the observer filter state space is reduced by removing an integral term in the observer, in which way the controller’s structure is further simplified while the asymptotic stability of the system is still maintained. Third, the control law is adaptive to various unknown system parameters including the mass of the vehicle, thrust coefficient, and bias errors in Euler angle measurement. Fourth, most traditional DIBVS control laws are developed relying on the measurement of the vehicle’s linear velocity. However, in GNSS-denied situation, this may not be practical. In this thesis, an output feedback method is adopted to DIBVS and the need of linear velocity measurement is removed.The controller guarantees the asymptotic stability of the error dynamics. Both simulation and experimental results are presented to demonstrate the performance of the proposed control law. The experiment architecture includes a quadrotor equipped with an open source Pixhawk autopilot and a CMUcam5 Pixy onboard computer vision system. The experimental result shows that the errors reach the steady state after about 20 seconds. At the steady state, the errors oscillate within a very small range where for position and heading errors, the oscillation amplitudes are less than 0:1 m and 0:1 rad, respectively. Master of Science (Aerospace Engineering)

Force is one of the major key parameters in a cutting process. The measurement of this parameter allows for assessment of the machining operation as well as to monitor tool condition (wear, breakage, etc.) and machine errors (looseness, backlash, etc.). The low cost option for measuring force in manufacturing is to use a load cell which is based on strain gauges. This technology has two drawbacks; firstly, strain gauges are not sensitive enough compared to other technologies and will not detect subtle changes in the cutting forces. Secondly, due to their design, load cells are mounted far from the cutting tool which means that the measured force will be not be representative of the actual cutting force. In the research environment, dynamometers are the most common technology that is used for measuring cutting forces in machine tools due to their high sensitivity. These dynamometers can also be mounted directly to the tool holders. They are based on four piezo-electric sensors which are preloaded between top and bottom plates like a sandwich. This dynamometer technology has not found its way into the production line, as it is relatively expensive (around 50,000 Singapore Dollars). Research is going on to develop a similar design of these dynamometers based on a single force sensor. The use of one sensor means that a calibration with a real dynamometer should be performed. In this project a force is applied using a robot arm on the customized dyno and then the readings recorded arc utilized to calibrate and compare them. In this research study, calibration is achieved using the least squares method and artificial neural networks followed by further analyses. Master of Science (Aerospace Engineering)

The composition and function of the intestinal microbiota plays an important role in human health and disease prevention. Prebiotics are promising health foods because they can regulate the structure and number of intestinal microbes. Prebiotics can stimulate the growth or activity of beneficial bacteria in the colon. Some Australian honeys have been shown to have prebiotic properties in adults and children with doses of 1% in microcosm experiments, but dose response studies are lacking. In this study, an Australian honey with demonstrated prebiotic properties, was used to study the dose response for infant microbiota. Ex-vivo microcosms seeded with fecal slurries from a 3 year old were established with different honey concentrations (1%, 0.75%,0.5%,0.25%,0). The microcosms were sampled after 48 h incubation and the samples analyzed for the short chain fatty acids (SCFAs), using GC-MS, and levels of the potentially beneficial Lactobacillus and Bifidobacterium by qPCR. The results show that different doses of honey have different effect on the infant’s microbiota. The honey concentration of 0.5% yielded the highest total number of Lactobacillus and Bifidobacterium. Furthermore, honey addition resulted in elevated levels of SCFAs, especially butyric acid, known to have beneficial effects in the body, however, with the higher concentration, lower levels of isovaleric acid and valeric acid were detected. It was concluded that lower doses of honey would sufficient for children, than are noted for adults in previous studies. ​Master of Science (Biomedical Engineering)

Quasi-Static loading analysis as an analogous to low-velocity impacts on the hybrid Fiber Metal Laminate (FML) is a conscious effort to characterize and study the damage mechanism associated with the one of the most common causes of failure in the hybrid material. With the latest furor in the field of FML structures: a lighter yet stronger alternative to conventional metals, extensive work has been carried out in analyzing quasi-static loading with regards to quantifying damage and predicting the onset of damage because of the barely visible impact damage. However variation in the constituent build-up of the material yields varied responses; the effect of the stacking sequence, the influence and contribution ofthe plasticity offered by the metallic element to the overall response, and the investigation of the damage behavior as a result of using different alloys of the same metal are some important aspects that require discemment. This report details the quasi-static loading analysis of FMLed circular plates and beams. A set of composite plate specimens with varying stacking sequences were fabricated and subjected to controlled quasi-static loading (lmm/min). FML plates and beams were fabricated with pre-calculated varying metal volume fractions (MVF) were also analyzed. Two sets of FML plates with two different aluminum alloys (AI 2024-T3 and 0) were also subjected to the quasi-static loading. Experiments revealed that a quasiisotropic stacking orientation exhibited enhanced load bearing and damage tolerance behavior than other stacking sequences. It was also realized that increasing MVF delayed the onset of damage, but lowers the load bearing capacity, and worsens the damage tolerance and resistance. It is leamed that Al2024-0, displays highly enhanced plasticity, but significantly lower load bearing capacity. It demonstrated better damage resistance than a similar FML plate with Al 2024-T3 alloy. A systematic finite element analysis (FEA) of all the above cases were performed using the commercial software, Abaqus. in order to study the progressive distribution of stress, explicitly analyze the failure associated with each damage mode and also to validate the experimental results. The numerical simulation was found to be in fairly good agreement with the experimental results in terms of force history and damage behavior. Master of Science (Aerospace Engineering)

Optical microscopy is often used to characterize the surface microstructure of polycrystalline materials because of its simplicity and low-cost. However, the type of information acquired is limited by the diffraction barrier of optical lenses. For this reason, optical microscopy cannot be employed to resolve materials’ atomic structure and its use is typically limited to quantifying the geometry and distribution of the microstructural features that can be resolved, such as crystal grains, phases, and defect. In this project we will use modified optical techniques and develop novel numerical analysis methods to overcome the diffraction barrier of optical microscopy and assess grain orientation information. Our techniques rely on the quantification of reflected light intensity as a function of incident light direction or light polarization. Our results lay the foundation for the development of new optical-based techniques for cost-effective and high-throughput microstructure characterization of polycrystalline materials. Master of Science (Aerospace Engineering)

The utilization of daylight can significantly affect building performance, energy efficiency, productivity, as well as occupants' comfort and satisfaction in buildings. This paper aims to assess daylight performance for tropical office buildings in a parametric approach. Thus, four passive design categories are investigated, namely interior surface reflectance, glazing visual transmittance, light shelves and shading control. The approach is exemplified using the case study of two selected offices in Singapore. This study contributed to the assessment of the daylight performance and prediction of the consequences of retrofitting alternatives toward fostering the utilization of daylight in existing buildings in the tropics. Published version

We have carried out a theoretical and numerical study of disorder-induced changes in the absorption line shape of a chromophore embedded in a host matrix. The stochastic sudden jump model is employed wherein the host matrix molecules are treated as noninteracting two-level systems (TLSs) occupying points on a three-dimensional lattice with randomly oriented dipole moments. By systematically controlling the degree of positional disorder (α) attributed to them, a perfectly crystalline (α = 0) or a glassy environment (α = 1) or a combination of the two is obtained. The interaction between the chromophore and the TLSs is assumed to be of the dipole–dipole form. With an increase in α, the broadening of the absorption line shape was found to follow a power-law behavior. More importantly, it is revealed in the long-time limit that the resultant line shape is Gaussian in the absence of disorder but transforms to Lorentzian for a completely disordered environment. For an arbitrarily intermediate value of α, the resultant line shape can be approximately fitted by a linear combination of Gaussian and Lorentzian components. The Lorentzian profile for the disordered medium is attributed to the chomophore–TLS pairs with vanishingly small separation between them. Accepted version

This paper presents analytical modeling of slip liquid flow in parallel-plate microchannels, and can be divided in two parts. In the first part, classical relationships describing velocity, flow rate, pressure gradient, and shear stress are extended to the more general cases where there exist two different values of the yet-unknown slip lengths at the top and bottom walls of the channel. These formulations can be used to experimentally determine the values of slip length on the channels fabricated from two different hydrophobic walls. In the second part, the emphasis is given on the quantification of the slip length analytically. Generating mechanism of slip is attributed to the existence of a low-viscosity region between the liquid and the solid surface. By extending the previous works, the analytical values of slip length are determined using exact, rather than empirical, values of air gap thickness at different ranges of air flow Knudsen number. In addition to the exact expressions of air gap thickness, the corresponding ranges of the channel height where slip flow can be induced are also found analytically. It is found that when the channel height is larger than 700 μ m, air flow is in continuum regime and no-slip boundary condition can be used. For the case where the channels height is smaller than 700 μ m, and larger than 7.5 μm, slip boundary condition should be used to model the air flow in the channel. Finally, for the channel with the height smaller than 7.5 μm, Navier-Stokes equation cannot be used to model the air flow, and instead molecularbased approaches should be implemented. The results of this paper can be used as a guideline for both experimentalists and theoreticians to study the slip flow in parallel-plate microchannels.

The optical properties of plasmonic nanoring pentamers are experimentally and theoretically investigated in the near-infrared spectral region, where the nanoring metamolecules allow complex optical design by tuning a single geometrical parameter (the inner radius of the nanoring). Even in fully four-fold symmetric nanoring pentamers, unconventional plasmonic coupling paths generate multiple Fano resonances which would not appear in symmetric nanodisc oligomers. In addition, inhomogeneous elementary units in the pentamer induce additional resonances due to high-order dark modes, while complete breaking of the four-fold symmetry results in sharp multipolar Fano resonances upon the hybridization of dark and bright modes. This approach demonstrates a simple way to engineer plasmonic couplings by subtle changes of uniform metamaterial design. ASTAR (Agency for Sci., Tech. and Research, S’pore)

This paper seeks to examine the concrete contribution to shear strength and determine the inclination of the compressive strut within the variable truss model for slender RC shear-critical beams with stirrups. Using the modified compression field theory in place of the conventional statistical regression of experimental data, the expression for the concrete contribution to shear strength was derived, and the inclination of compressive struts was determined. A simplified explicit expression for shear strength was then provided, with which shear strength can be calculated without extensive iterative computations. This method was then verified using the available experimental data of 209 RC rectangular beams with stirrups and compared with the current methods from the American Concrete Institute and the Canadian Standards Association. The theoretical results are shown to be consistent with the experimentally observed behavior of shear-critical RC beams.

Experimental observations on 11 Reinforced Concrete (RC) columns tested at Nanyang Technological University (NTU), and existing experimental data of 79 shear-critical RC columns are presented. Significant arch action is found in columns of small shear span-to-depth ratio and high axial-load ratio under shear force. Utilizing the sectional method for shear strength of these types of columns, which does not consider arch action, would give a more conservative prediction. Based on the truss-arch model, an expression to predict the shear strength of shear-critical RC columns is presented, which considers both the contributions of concrete and transverse reinforcement to shear strength in the truss model as well as the contribution of arch action through compatibility of deformation. The proposed model is compared with other shear strength models using the available column test data consisting of 90 shear-critical RC columns, and the results show that the proposed model can improve the accuracy of shear strength predictions for shear-critical RC columns Accepted version

Sessile nanofluid droplet evaporation is receiving increasing attention lately due to its potential in various applications requiring controlling of the solutes morphology after liquid drying. Despite extensive progress, the understanding still remains elusive about how the deposited morphology is affected and produced owing to the numerous interrelated factors of a nanofluid droplet system. I investigate sessile nanofluid droplet drying in respect to the domains of colloidal suspensions, base solutions and solid substrates. In respect to suspensions, the addition of mono graphite nanoparticles is found to reduce the droplet wettability, enhance the contact line pinning effect, and induce a slight spreading at the initial stage of evaporation. By using proper combinations of dual-sized alumina nanoparticles, the heterogeneities of the two single-sized nanoparticle patterns are eliminated, and a uniform mixture pattern is produced. Besides, the mixture pattern assembled by two nanoparticle species either exhibits both the features of the single-species patterns or resembles one of them. In respect to the base solution, we probe the flow paradigms during evaporation from two aspects. One is employing water-ethanol binary droplets to examine the effect of ethanol concentration. The binary droplets display three distinct flow regimes. The relative weighings of Regimes I and II are enhanced and Regime III is shortened upon raising the ethanol component. The ethanol dependent weighings of the three regimes determine the congregation and trajectories of the nanoparticles and thus the final deposition. The other study for base solution is probing the autophobic effect induced by cationic surfactant cetyltrimethylammonium bromide in sessile droplets. For the pure droplet, below the critical micelle concentration of the surfactant, the droplet dominated by the autophobic effect exhibits two distinct phases of depinning: Phase 1 featuring rapid droplet shrinkage, and Phase 2 characterized by slower droplet receding. The velocity of the three-phase contact line in Phase 1 shows a transition as the surfactant concentration increases above 0.042 mM, while such a transition is absent for Phase 2. Besides, the spreading of the sessile droplets as they form before the retraction and the maximum contact angle led by dewetting are found regularly dependent on the surfactant concentration. For graphite nanofluid droplets, the autophobic effect is enhanced upon increasing the nanoparticle concentration. The deposited morphology evolves to a distinct coffee ring with a loading of the surfactant, owing to the surfactant adsorption at the nanoparticles surface and the liquid-solid interface, which repels the nanoparticles from the liquid-vapor and the liquid-solid interfaces. In respect to the solid surface, the substrate temperature was varied to find its role in controlling the flow patterns and the stain morphology. The deposited patterns transforms from a disk-like profile to a dual ring from cooling to heating of the substrate. The droplet on the substrate at low a temperature reveals three primary stages. Stage I features the outward transports of nanoparticles along the liquid-vapor interface near the droplet edge. Meanwhile some nanoparticles deposit on the solid surface with a distance to the contact line. In the central region, nanoparticles are dominated by Brownian motion, so they fluctuate irregularly around their positions. Stage II is characterized by the enhanced outward travellings of the nanoparticles in the bulk, leading to a pronounced coffee ring. Most nanoparticles in Stages I and II are central-concentrated, leaving an annular gap sparsely covered adjacent to the outer ring. In Stage III, the pattern is homogenized by filling the gap of the interior nanoparticles. Upon increasing the substrate temperature, the accompanied flow pattern displays a transition when the substrate is still remained cooler than the atmosphere. It is attributed to the evaporative cooling at the droplet apex counteractive to the applied temperature gradient by substrate cooling. Above the transition temperature, the induced inward Marangoni flow takes place earlier at a higher substrate temperature, and in conjunction with the outward radial flow, a dual ring pattern is formed. Recommended future works comprise the employments of dual-sized fluorescent spheres to examine the size-dependent motion of spheres, textured substrates for distinct droplet dynamics, and the infrared thermographic technique for detecting the temperature profile of droplet free surface. Doctor of Philosophy (MAE)

The ever increasing digital data require memory and logic devices that have high speed, need low power, offer low cost and are reliable. Domain wall (DW) based memory and logic devices potentially offers all the above advantages along with being non-volatile. In this thesis, the dynamics of DWs have been investigated in ferromagnetic nanostructures for realizing efficient logic and memory based applications. The study has been divided into two parts. In part one, the DW dynamics in ferromagnetic network structures with in-plane magnetic anisotropy (IMA) is studied. The magnetic material chosen is NiFe due to its large magnetic permeability and near zero crystalline anisotropy. In part two of the study, materials with perpendicular magnetic anisotropy (PMA) are considered. Here, Co/Ni and Co/Pt multilayers are chosen as the system for study. Field induced domain wall driving in IMA network structures revealed the DW trajectory to be chirality dependent. Deterministic trajectory of DW in the network structure is demonstrated for path lengths less than the fidelity length and for fields less than Walker breakdown field. For distances in excess of fidelity length, Walker breakdown is observed where DW chirality is found to oscillate and the trajectory becomes stochastic. To overcome this issue, the network structures are re-designed to include geometrical asymmetry. This constrains DW to propagate in a definite and fixed branch, by raising the potential barrier in the other branch, irrespective of its chirality. This branch engineering is applied to propose and demonstrate a programmable logic device. The control of the trajectory is provided by fabricating a metallic gate on top of the network structure. The current in the gate applies an Oersted field which influences the trajectory of DW by interacting with the transverse charge distribution in the DW. A transverse nanowire in the structure serves a dual purpose of acting as an input logic and also transforming the vortex DW to a transverse chirality. Two universal logic gate functionalities NAND and NOR are demonstrated along with the complimentary logic operations in the same structure using magnetic force microscopy imaging and anisotropic magneto-resistance. The PMA system is investigated using the phenomena of anomalous Hall effect. Our investigations reveal multiple DW nucleations in low anisotropy Co/Ni Hall cross junctions under the application of in-plane current without the need of conventionally used local Oersted field. The stochasticity in the nucleation process due to high demagnetization energy at the Hall cross coupled with joule heating is utilized to demonstrate a DW based random number generator device. The analog signal from the device is fed to an integrated circuit to amplify and digitize the output. Recent developments have shown combination of interfacial phenomenon such as spin-orbit torque (SOT) and Dzyloshinskii-Morya interaction leading to high speed DW dynamics. Pt/Co/Ta stacks have enhanced SOT strength due to opposite signs of spin Hall angle of Ta and Pt. However, it is observed that stack possesses low thermal stability due to reported intermixing of Ta and Co. A solution to this issue is provided by adding additional Co/Pt interface which enhances the thermal stability and PMA strength. The distribution in velocity with the applied current density and spin Hall angle is estimated by changing the Ta capping thickness and Pt spacer thickness. SOT driven DW speed of about 530 m/s at a current density of 1×1012 A/m2 is demonstrated. This high speed is attributed to low anisotropy of our device coupled with in-plane field applied to prevent Walker breakdown. DW velocity was higher and along the current flow direction when the in-plane field assisted the inherent DMI field, while it was lower and along the electron flow direction when the field opposed the DMI field. Doctor of Philosophy (SPMS)

Hybrid superplastic forming (SPF) has recently been developed as a sheet forming technique, combining a hot drawing (mechanical pre-forming) process and a gas forming process. The hot drawing stage greatly enhances the formability and the forming efficiency of the metal sheet. The blank sheet with a desired amount of material is drawn into the die to form an intermediate hollow shape by a punch. During the gas forming stage, the mechanical pre-formed part is further formed at a target strain rate by applying the gas pressure. It is considered as an efficient way to form components with complex shapes in automotive and aerospace industries. In this work, non-superplastic grade AZ31B magnesium sheets were successfully formed by hybrid SPF at 400 °C. The dome height of formed part was 60 mm under gas pressure cycle S1, approximately 60 % of the die diameter. In contrast, the part formed by conventional SPF cracked at the same height. The influence of punch shape on the formability of AZ31B was also investigated. The minimum thickness of 1.31 mm after hybrid SPF was found at location where the material first came into contact with the punch, and the corresponding thickness reduction was 59%. Compared with the conventional SPF process, the thickness distribution of part formed by hybrid SPF was significantly improved. Additionally, the microstructure evolution of AZ31B in uniaxial tensile tests and hybrid SPF were examined by electron backscatter diffraction. The static and dynamic grain growth at grip region and gauge region were also investigated at various strain levels. Many subgrains with low misorientation angle were observed in the coarse grains during SPF process. The average misorientation angle and fraction of high angle grain boundaries decreased with increasing strain. This explains well the generation of subgrain via dislocation motion during SPF. Therefore, the main deformation mechanisms during hybrid SPF was determined to be a combination of recrystallization, dislocation climb and grain boundary sliding (GBS). Based on the tensile test results, parameters of hyperbolic sine creep law model and two-term material constitutive model were determined at 400 ºC. The hybrid SPF behavior of non-superplastic grade AZ31B was predicted by ABAQUS using these two material forming models. The FEM results of thickness distribution, thinning characteristics and forming height were compared with the experimental results. For the hot drawing process, the simulation results of both two models agree well with the experimental results. However, neither the hyperbolic sine creep law material model nor the two-term material model could predict the dome height and thickness distribution well during hybrid SPF. There is an increase in deviation between the prediction and experimental results when the material undergoes larger strain deformation during SPF. As a result, the part formed by hybrid SPF was done faster with higher dome height and a more even thickness distribution than conventional SPF. The forming time for hybrid SPF was significantly shortened as the hot drawing step formed the pre-formed component rapidly before the final more time-consuming gas blow forming. Accordingly, AZ31B part formed by hybrid SPF is structurally stronger due to the presence of fewer and smaller cavities as compared to conventional SPF. Doctor of Philosophy (MAE)

Instrumented projectiles are widely used in the arena of military, geological studies, planetary and comet mission. Their applications in harsh environments call for a rugged and highly reliable system to ensure the survival of the internal electronics and reliable measurement of in situ data. These data are essential to study the various properties of the penetrated target and the projectile. The design, development and construction of a novel instrumented projectile for in situ deceleration measurement in high velocity impact conditions are presented in this thesis. A series of impact tests were performed using the developed instrumented projectile. The obtained acceleration responses from the series of tests, composed of transient, vibration and rigid-body responses were studied using frequency and frequency-time analysis to deduce the dominant resonance frequency of the instrumented projectile. Based on this deduction, optimum parameters for the design of the digital low-pass filtering schemes are obtained, which subsequently provides a consistent framework for the remaining studies in this work to extract the rigid-body deceleration of the instrumented projectile. The rigid body deceleration profile is important for examining the resistance of targets and verifying various prediction models. The novel instrumented projectile was employed to study the resistance characteristics of closed-cell aluminium foams of two different density groups and gauge lengths. The study shows that inertia effect has a larger influence on the rate sensitivity of the low density foams compared with the high density foams. This study also shows that the strain hardening behaviour of foam specimens and the normalised penetration depth are important factors that dictate the deceleration response profile of the projectile. The abovementioned factors are important to design an efficient energy absorber that controls the deceleration profile of a colliding mass. An analytical model is proposed to predict the responses of instrumented projectile when it is impacting on closed-cell aluminium foams. Predictions by the models are compared with experimental data and with the results obtained from three-dimensional finite element simulations. This proposed model is based on an exponential stress-strain relation that takes into consideration of the strain-hardening behaviour of the aluminium foam specimens. It was found that the proposed model gives a better prediction than another earlier model that is based on a Rigid Perfectly-Plastic Locking idealisation that dictates the foam’s resistance, which is only able to approximate the stress-strain characteristic of metal foams with low strain hardening. Doctor of Philosophy (MAE)

Variability is the greatest enemy of a manufacturing system, which causes unstable performances and varies the productivity. This report focuses on reducing the variability and improving the performance of a semiconductor manufacturing process. Other methods to improve the performance of a semiconductor manufacturing process includes improving machine's process time, standard work of operators and other ergonomics factors; these are included in this study. A simulation model of the manufacturing semiconductor cell was created using the Arena software. It was based on actual production process data from the author's company Heptagon. To get different performances of the manufacturing system, some scenarios of variations had been applied on the simulation model. The results were used in further analysis and discussion to investigate the factors that produce the best improvement or cause the worst productivity in the manufacturing system. The author proposed a set of solutions to improve the performance of the semiconductor manufacturing system. These include improvements on number of operators, walking speed of operators, process analysis time reduction and machine modification time reduction. Master of Science (Manufacturing Systems & Engineering)

Customer experience IS a crucial aspect of product or service design. A good customer experience means the great satisfaction of the customer expectation. In the service industry, conversely, the customer experience failure of service performance means the service failure. Service recovery as a measure of service failure handling requires more and more complex failure analysis and customer research. Service failure must spend the massive cost for processing corresponding recovery measures according to a specific failure situation. Attempting to enhance the ability of service recovery and customer experience of Expo (Exposition) process, this research proposes a novel systematic service design approach named "Dynamic Experience TRIZ service design model (DETRIZ model)" that incorporates TRIZ (Inventive Problem Solving Theory), Customer Experience Journey Map (CXJM), service assembly concept and service replace concept. Then, new service systems named smart Expo service were designed for case companies, in order to verify the proposed service design approaches. This dissertation applies the user-centered method, Customer Experience Journey Map. It analyses the customer pain points during the three basic MICE (Meetings, Incentives, Conventions, and Exhibition industries) industries, Exhibition, Tourism, and Hotel. Then, the innovation problem-solving theory (TRIZ) is introduced and utilized to deal with the service provider's pain points. Based on TRIZ innovative principles, the Service Assembly Concept is represented as a resolution to enhancing the customer experience. The service assembly concept cooperating with emotional range analyses, and then design two service assembly mechanisms. The service assembly mechanism improves the overall customer experience. Besides, the service replacement mechanism is a simplest service failure handling measure. A case study of a Kaohsiung exhibition center in Taiwan is taken as the validation. The final empirical result shows the improvements of customer experience satisfaction and effectiveness of failure handling. Master of Science (Smart Product Design)

An analytical approach based on modified Bouc–Wen–Baber–Noori model has been proposed in this paper for predicting the hysteresis behavior of reinforced concrete beam–column joints with limited transverse reinforcement. The analytical model presented in this research is able to capture the characteristics of non-seismic detailed beam–column joints such as stiffness and strength degradation and pinching. Livermore Solver for Ordinary Differential Equations (LSODE) and Genetic Algorithm (GA) have been employed to solve the differential equations and to execute systematic estimation of the parameters associated with the model respectively. The analytical model has been calibrated with the experimental results of old fashioned interior and exterior beam–column joints obtained from the literature. In a bid to examine the influence of variation of each analytical parameter on the model, sensitivity analysis has been performed. Thereafter, an extensive parametric study has been conducted to relate the physical parameters of beam–column joints to the analytical model parameters. The upper and lower bounds of the magnitude of the analytical model parameters have been proposed subsequently with a method to identify the parameters for a specific beam–column joint depending on its physical parameters. Accepted version

Chemical looping combustion (CLC) of coal has received great attention for its verified advantage in the inherent separation of CO 2 without great cost penalty. It would be meaningful to adopt anthracite as fuel for CLC application in China due to its abundant reserve in China and also a great challenge for its low reactivity. A typical Chinese anthracite as Yang Quan (YQ) coal was selected in this research as a model fuel and its reaction with both synthesized CuO/Al 2 O 3 and Fe 2 O 3 /Al 2 O 3 oxygen carriers (OCs) was performed in a thermogravimetric analyzer (TGA) to investigate the reaction peculiarities involved. Fourier transform infrared spectroscopy (FTIR) was used to in-situ detect the emitted gasses from TGA. Field scanning electron microscopy/energy-dispersive X-ray spectrometry (FSEM-EDX) was used to study the morphology and elemental compositions present in the solid residues collected from the reaction of YQ with these two OCs, and the related phase was identified by X-ray diffraction (XRD). In order to further explore the reaction mechanisms involved, a more realistic reaction system with 426 species was designed for thermodynamic simulation. Through all these measures, two reaction stages were observed for the reaction of YQ with Fe 2 O 3 /Al 2 O 3 or CuO/Al 2 O 3 at 300–600 °C and 600–850 °C after dehydration, respectively. The maximum weight loss rate for YQ with CuO/Al 2 O 3 at the second stage was pronounced enough to reach up to 2.8 wt.%/min, greatly higher than that of YQ with Fe 2 O 3 /Al 2 O 3 . The mixture conversion index for YQ with CuO/Al 2 O 3 was far bigger than that of YQ with Fe 2 O 3 /Al 2 O 3, which further indicated that CuO/Al 2 O 3 was more suitable to YQ in CLC. At these two stages throughout 400 to 1100 °C, CuO and CuAl 2 O 4 contained in CuO/Al 2 O 3 OC were mainly reduced to Cu and Cu 2 O or CuAlO 2 and Al 2 O 3 by transfer of the lattice oxygen [O] involved to YQ in a sequential mode; however, above 800 °C, Cu 2 O and CuAlO 2 were also produced through direct decomposition of CuO or CuAl 2 O 4 by emission of gaseous oxygen O 2 , which was beneficial to the direct combustion of coal. During reaction of YQ with CuO/Al 2 O 3 OC, active CuO was found to tend to react with various sulfur species produced from YQ pyrolysis and formed to Cu 2 S, but the Al 2 SiO 5 was formed through the interaction of Al 2 O 3 with SiO 2 in the YQ, which resulted in the loss of the inert support involved in the CuO/Al 2 O 3 OC and further degraded its reactivity to YQ due to the lower resistance to sintering.

This paper attempts to determine the inclination of the compression strut within variable angle truss models for RC beams loaded in shear-flexure through a proposed semi-analytical approach. A truss unit is used to analyze a reinforced concrete beam, by the principle of virtual work under the truss analogy. The inclination of the compression strut is then theoretically derived. The concrete contribution is addressed by utilizing the compatibility condition within each truss unit. Comparisons are made between the predicted and published experimental results of the seventy one RC beams with respect to the shear strength and the inclined angle of the compression strut at this state to investigate the adequacy of the proposed semi-analytical approach. Published Version

The estimation of the initial stiffness of columns subjected to seismic loadings has long been a matter of considerable uncertainty. This paper reports a study that is devoted to addressing this uncertainty by developing a rational method to determine the initial stiffness of RC columns when subjected to seismic loads. A comprehensive parametric study based on a proposed method is initially carried out to investigate the influences of several critical parameters. A simple equation is then proposed to estimate the initial stiffness of RC columns. The applicability and accuracy of the proposed method and equation are then verified with the experimental data obtained from literature studies. Accepted version

Columns are the key load-bearing elements in frame structures and exterior columns are probably the most vulnerable structural components to terrorist attacks. Column failure is normally the primary cause of progressive failure in frame structures. A high-fidelity physics-based computer program, LS-DYNA was utilized in this study to provide numerical simulations of the dynamic responses and residual axial strength of reinforced concrete columns subjected to short standoff blast conditions. The finite element (FE) model is discussed in detail and verified through correlated experimental studies. An extensive parametric study was carried out on a series of 12 columns to investigate the effects of transverse reinforcement ratio, axial load ratio, longitudinal reinforcement ratio, and column aspect ratio. These various parameters were incorporated into a proposed formula, capable of estimating the residual axial capacity ratio based on the mid-height displacement to height ratios. Accepted version

The selection of controlled variables (CVs) from available measurements through exhaustive search is computationally forbidding for large-scale processes. We have recently proposed novel bidirectional branch and bound (B3) approaches for CV selection using the minimum singular value (MSV) rule and the local worst-case loss criterion in the framework of self-optimizing control. However, the MSV rule is approximate and worst-case scenario may not occur frequently in practice. Thus, CV selection by minimizing local average loss can be deemed as most reliable. In this work, the B3 approach is extended to CV selection based on local average loss metric. Lower bounds on local average loss and, fast pruning and branching algorithms are derived for the efficient B3 algorithm. Random matrices and binary distillation column case study are used to demonstrate the computational efficiency of the proposed method. Published version

This study investigates the applicability and reliability of the strut-and-tie modelling that is recommended by the ACI 318 Building Code (2002) for designing continuous deep beams with openings. Presented herein are experimental results of two full-scale continuous reinforced concrete deep beams with openings subjected to concentrated loads. Each specimen was designed using the ACI 318 (2002) provisions for strut-and-tie models. Both specimens were found to be able to carry loads greater than the factored design load. The failure mechanism of each specimen corresponds well with the designed strut-and-tie models. The experimental results thus illustrate the reliability and conservative nature of strut-and-tie modelling for designing continuous deep beams with openings.

Three nonseismically detailed interior reinforced concrete beam-column joints, namely, one eccentric and two concentric joints, strengthened with proposed fiber-reinforced polymer FRP wrapping configurations using glass fiber-reinforced polymer and carbon fiber-reinforced polymer strips and sheets, were tested under constant axial compression load and reversed cyclic loading which simulated low to moderate earthquake forces. Seismic performance of the strengthened beam-column joints in terms of their hysteresis response, stiffness, and energy dissipation capacity is evaluated and compared to those of the original and unstrengthened beam-column joints. Results indicate that applying strips at 45º on a flushed eccentric joint core and as cross bracing on the beam and confinement round the column is very effective. All specimens failed with gradual strength deterioration, bond degradation, and debonding of FRP sheets was observed near the joint core. The proposed strengthening schemes were found to be efficient and economical for mass repair or upgrading of nonseismically detailed structures. Accepted version

This paper examines the impact responses of reinforced concrete RC beams through an experimental study and presents an analytical model developed to predict the maximum midspan deflection and maximum impact load, which aids as an important performance index to evaluate the damage levels of RC beams when subjected to impact loadings. The experimental study involves a drop hammer impact test and investigates the influence of drop height and the effect of the amount of longitudinal steel reinforcement contributes to the response of RC beams. The RC beam specimens used in the experiment comprised of under-reinforced sections provided with sufficient amount of transverse reinforcements to allow for an overall flexural failure. The experimental impact responses of the RC beams were simulated with two-degree-of-freedom mass-spring-damper system model, in which the loading rate effects were duly considered. The analytical results are in good agreement with the experimental results for the RC beams that exhibited overall flexural failure. Accepted version

In regions of low to moderate seismic risk, such as Singapore and Malaysia, buildings with relatively weak lateral structural resisting systems are likely to be common. Although ground motions because of long distance earthquakes centred in Sumatra have occurred, there has been no record of earthquake damage in this region. This article compares the results obtained from pushover analyses of six-storey reinforced concrete (RC) frames with experimental observations. The experimental observations of four internal beam-column joints, which were designed according to BS 8110, were summarised. From this, we address the performance of some typical RC frame structures under low seismic loading. By adding about 0.075% of hoop reinforcement in the beam-column joints, one can improve the seismic performance of the frame in the weak direction more significantly than in the strong direction. Also, relatively large joint shear input during low to moderate earthquakes can result in diagonal tension cracking and the building may likely be governed by joint shear failure.

In this article, experimental and finite element (FE) numerical investigations on interior wide-beam column joints are presented. The experimental research consisting of three full-scale interior widebeam column specimens was carried out at Nanyang Technological University, Singapore to study the seismic behavior. Details of the test results are discussed to understand the specimens’ seismic performance in terms of general behavior, hysteresis loops response, and strain profiles of longitudinal reinforcement. In the FE numerical study, the three-dimensional (3D) model developed is validated by comparing the analysis results with the experimental test results, which has shown a good agreement. A parametric study is performed to elucidate more information and to understand the influence of critical parameters affecting the joint behavior such as column axial load, beam anchorage ratio, and wide beam participation. Accepted version

A truss model is presented in this paper to predict the load deflection response of reinforced concrete beams subjected to flexure and shear. The truss model which has struts at various angles is derived considering concrete contribution. The concrete contribution is then integrated into the modified truss model through the concept of equivalent stirrup reinforcement. The validity and applicability of the proposed truss model is evaluated by comparing results with experimental data. Good correlations with the experimental data have been obtained. This study shows that the truss model analogy, if properly treated, can be used to access both shear strength as well as the load deflection response of reinforced concrete elements subjected to shear and flexure. Accepted version

Both maximum displacement and displacement ductility factors should be considered in the design of a blast-resistant structure since both parameters correlate with an expected performance level of a reinforced concrete RC structural member during a blast event. The blast-resistant design procedure discussed in this paper takes into account both the maximum displacement and displacement ductility responses of an equivalent single-degree-of-freedom SDOF system, while the response of the SDOF system is made equivalent to the corresponding targets of design performance. Some approximate errors are present when comparing the actual responses of the structural member, which has been designed for blast loading, and their corresponding design performance targets. Two indices are defined to quantify the approximation errors, and their expressions are obtained through comprehensive numerical and statistical analyses. By using the error indices, the design procedure is then modified such that the approximate responses of the RC member are equivalent to the targets of the design performance. The modified procedure is implemented in three design examples and numerically evaluated. It is concluded that the modified procedure can be used more effectively in order to ensure that the actual responses of designed members reflect the respective targets of design performance. Accepted version

In this paper, a nonlinear finite element (FE) analysis of hybrid-steel concrete connections is presented. The detailed experimental results of the four full-scale hybrid-steel concrete connections with limited seismic detailing have been discussed in a different paper. However, due to the inherent complexity of beam–column joints and the unique features of the tested specimens, the experimental study was not comprehensive enough. Therefore, in this paper, an analytical investigation based on the FE models and using the DIANA software is presented. The FE models were validated using the experimental results of the hybrid-steel concrete connections tested in Nanyang Technological University, Singapore. The critical parameters influencing the joint’s behaviour, such as the axial load on column, the connection plate thickness, and the continuation of beam bottom reinforcement, are varied, and their effects, especially implications on code specifications, are studied. Accepted version

This thesis proposed a novel ejector based multi-evaporator refrigeration system (MERS). The energy inefficiency of conventional multi-evaporator refrigeration system was analyzed and this novel ejector based multi-evaporator refrigeration system was introduced. The working principle is described in detail and the possible advantage and disadvantage is discussed. The experimental platform is built to conduct the comparison study between conventional MERS and ejector based MERS. The system is introduced from architecture to detailed functional programming. Experiments have been done on the experimental platform to compare the energy consumption of these two systems. By analyzing the performance of the novel ejector based multi-evaporator refrigeration system, several improvements and future works are proposed. Master of Engineering

Multi-channel signal observations in biomedical, radar and other communication applications are multivariate and contain contributions from multiple sources. Independent Component Analysis (ICA) has been widely used for separation of underlying sources from the observed mixtures for the subsequent analysis of such data. To e ciently extract a subset of desired sources, the Constrained ICA (cICA) that provides a framework to incorporate the prior information about these sources as constraints into the contrast function of ICA was introduced. ICA with Reference (ICA-R) proposed under the cICA framework incorporates priori information as reference signals (rough template) that guide source separation to automatically extract desired sources closer in some sense to the reference signal. However, the convergence of cICA and ICA-R to the desired subset of sources as well as the consistency of the separations estimated, greatly determines its applicability to the real world problems of source extraction. The factors that greatly a ect the convergence are: 1. the non-smooth penalized contrast function of ICA-R 2. the closeness function and lack of knowledge to choose the upper bound for the closeness measure 3. the sensitivity of the Newton-like learning (optimization) algorithm to the initialization of parameters and its inherent limitation to achieve global optimum Although di fferent closeness measures and initialization strategies have been investigated to improve the performance of ICA-R, not much research on the improvement of the optimization process of ICA-R has been attempted. This thesis mainly focuses on improving the convergence and stability of ICA-R and proposes ways to overcome the drawbacks of Newton-like learning algorithm by taking the advantage of the population-based evolutionary approaches that asymptotically converge to the desired global optima. First, we investigate the pros and cons of ICA-R and then discuss about the recent works that address the convergence and stability of ICA-R. The major contributions of the thesis are as follows: (i) a detailed analysis on convergence and stability of ICA-R and new ways of obtaining reliable reference signals, (ii) to avoid the formulation of a penalized objective function, a constrained di fferential evolutionary algorithm, termed cDE-ICAR, with a specifi c initialization strategy that incorporates the violation of closeness constraint to the selection process of evolution is proposed, (iii) to completely remove the upper-bound of closeness measure from the optimization problem of ICA-R, we propose a multi-objective optimization algorithm, termed MO-ICAR, that simultaneously optimizes the contrast function and the closeness measure and (iv) the utility and efficacy of the proposed methods in practical applications was demonstrated by extracting desired rhythms and removal of artifacts from EEG signals. Doctor of Philosophy (SCE)

A fuel cell, which directly converts chemical energy into electricity has been a promising candidate as a next-generation renewable energy source. Especially, solid oxide fuel cells (SOFCs) have been recognized as high energy conversion efficient and fuel flexible devices. However, conventional SOFCs are typically operated above 700°C to activate ionic transport across an ionic conducting electrolyte. Such high operating temperatures limit to expand the scope of their applications and require special materials for sealing and heat management which is expensive and cause design challenges. Therefore, lowering operating temperature is a primary concern to expand the field of feasible application. Various efforts have been made to reduce operating temperature either by thinner electrolyte or by better catalytic materials. In particular, thin film electrolytes significantly reduce ohmic resistance of SOFCs and accordingly, provides good fuel cell performances at a low operating temperature regime (300 to 500°C). Such thin film can be fabricated as a free-standing membrane by the help of silicon micromachining technology. However, in order to minimize the internal resistance of the cell, the membrane is usually made to be only a few tens of nanometers in thickness, and subsequently its membrane stability is deteriorated. In addition, the lateral dimension of a fuel cell determines the total power output of this device, and enlarging the electrolyte to obtain higher power output without membrane failures becomes nearly impossible. In this dissertation, new fabrication methods to realize thin film SOFCs with enhanced membrane mechanical stability are presented. Different from conventional square thin films having geometrical discontinuities such as sharp corners, circular thin films have uniform stress distribution and no stress concentration points along the boundary due to axisymmetric geometry. By taking mechanical advantages of a circular thin film, we successfully fabricated millimeter-scale and 100 nm-thick circular yttria-stabilized zirconia (YSZ) and yttria-stabilized barium zirconate (BYZ) electrolyte SOFCs with a high yield rate above 80% by combination of silicon wet etching and plasma etching. Circular electrolytes were sustained by a silicon-edge support which acts as a stress absorber. Compared with the clamped square membranes, the circular membrane reinforced with a silicon-edge support showed 40% reduction in the maximum principal stress. Moreover, electrochemical performances of the circular electrolyte membranes showed that high open circuit voltage (OCV) above 1.07 V and maximum power densities of 437 mW/cm2 and 76 mW/cm2 at 400°C of the 1.5 mm diametral YSZ cell and 1.6 mm diametral BYZ cell, respectively. To further enlarge electrochemically active area with maintaining mechanical stability, a fabrication method of circular array SOFCs applied with silicon-edge reinforcement is demonstrated. In this work, we fabricated a new thin film SOFC architecture that can provide upward membrane scalability to obtain higher device power, and at the same time enhance membrane mechanical stability for stable fuel cell operation. This SOFC array achieved high OCV of 1.1 V and provided 1.38 mW of total power output at 400 °C with the array of 3.6 mm in lateral dimension. Through long-term OCV test and thermal cycling test, functional and thermal stability of the circular array SOFCs with edge-reinforcement were confirmed. With further design and process optimization, higher total power output can be achieved at low temperature by this μ-SOFC technology. Doctor of Philosophy (MAE)

Much attention has been paid on lithium-ion batteries recent years because of their wide applications in portable electronic devices as well as in vehicles, ships and aircrafts. However, lithium-ion battery has problems of poor safety characteristics caused by high heat generation and the temperature rise during charge and discharge. High temperature affects the performance, reliability and lifespan of the battery and may cause serious safety hazards such as thermal runaway, even ignition. Therefore, the investigation and thermal management of lithium-ion battery are critical issues in research field. This thesis is concerned with thermal investigation and management of lithium-ion battery and battery module. The key aspects in the exploration, design, analysis and management of lithium-ion battery/battery module covered in this thesis are in the sequence as follows: (1) simplified one-dimensional mathematical thermal model of lithium-ion single cell; (2) three-dimensional transient thermal model of lithium-ion single cell based on the actual structure of battery; (3) investigation of forced air cooling strategy and design of battery module structure for lithium-ion battery module based on empirical heat source model; (4) exploration of optimal forced air cooling method; (5) simulation on battery module liquid cooling method. Thermal models are essential part in the research and application of lithium-ion battery. In this thesis, both one-dimensional and three-dimensional thermal models are developed for different application conditions. The one-dimensional thermal model is simplified from existed electrochemical model which describes the reaction process of lithium-ion battery during charge/discharge in a mathematical approach to describe the battery temperature behavior during operation. The three-dimensional transient thermal model is based on the experimental data and the actual structure of lithium-ion battery. Thermal behaviors which include temperature changing trend and temperature distribution of battery under different working conditions are simulated. This thesis has also defined an empirical heat source model for battery module/pack simulation. Design, analysis and optimization of battery module with forced air cooling strategies as well as liquid cooling methods have been presented in this work based on empirical heat source model. The thermal behaviors of battery module under different cooling strategies and working conditions are simulated and compared, optimal cooling strategy are selected for applications. Doctor of Philosophy (EEE)