<p> This research was motivated by the need for functional change of complex systems in post-production. Through an initial empirical study, this research found pathways to functional change with minimal change to form; thus avoiding the high risk of material change propagation that has been a concern of both the systems engineering and engineering change literature. This study also revealed the relative importance of system users in the post-production change environment. Through a follow-on study of user innovation in this environment this research was able to reveal pathways for change which are underexplored in both the user innovation and flexibility literature. </p><p> While not yet a final solution to the motivating problem, this research has revealed levers for change which system users are adept at exploiting. Key amongst the changes employed by user designers is their unconstrained use of themselves (human change) to realize system function and their exploitation of the expansive operational states (operational change) afforded by complex systems. User driven change poses an important source of post-production functional change. This research unpacks user change and sets the stage for further investigation of this source of system changeability. </p>

<p> Home automation systems have gained popularity in recent years, paralleling the advances in the concept of the Internet of Things. The current project presents the implementation of an inexpensive home automation system, within the framework of assistive technology. The system implementation is based on the Arduino microcontroller, with Bluetooth communications capability, and it is designed for use by the elderly and people with disabilities. The system is user-friendly, with an intuitive interface implemented on an Android-based smart phone. Demonstrations show that the system facilitates control of home appliances, lights, heating, cooling systems and security devices by the intended users, i.e. the elderly and the disabled.</p>

More than 80% of the people in urban regions and about 98% of cities in low and middle income countries have poor air quality according to the World Health Organization. People living in such environment suffer from many disorders like a headache, shortness of breath or even the worst diseases like lung cancer, asthma etc. The main objective of the thesis is to create awareness about the air quality and the factors that are causing air pollution to the people which is really important and provide tools at their convenience to measure and analyze the air quality. Taking real time air quality scenarios, various experiments were made using efficient sensors to study both the indoor and outdoor air quality. These experimental results will eventually help people to understand air quality better. An outdoor air quality data measurement system is developed in this research using Python programming to provide people an opportunity to retrieve and manage the air quality data and get the concentrations of the leading pollutants. The entire designing of the program is made to run with the help of a graphical user interface tool for the user, as user convenience is considered as one of the objectives of the thesis. A graphical user interface is made for the user convenience to visualize graphically the data from the database. The designed system is tested and used for the measurement and analysis of the outdoor air quality. This data will be available in the database so it can be used for analyzing the air quality data for several days or months or years. Using the GrayWolf system and the designed outdoor air quality data measurement system, both the indoor and outdoor air quality was measured to analyze and correlate.

Electronic noise is an area of fundamental importance that reveals information about the system that transport or optical measurements cannot. The present work focuses on recent advances made in the area of noise in nanodevices. This area is attracting increasing attention because noise in nanostructures is qualitatively and quantitatively different from that in bulk systems. Chapter 1 discusses different kinds of noise in nanostructures with a focus on 1/f noise. Monte-Carlo simulation is used to find velocity as a function of time to calculate velocity autocorrelation function and noise spectral density. The spectrum has 1/f^2 dependence which is universal irrespective of electric field and temperature which implies that it is a fundamental effect. In analogy with electronic noise spin noise is studied in Chapter 2. Fluctuation in spin polarization of carriers in semiconductor structures gives rise to spin noise. Spin transport is simulated using Monte-Carlo simulator in presence of D’yakonov-Perel’ spin relaxation mechanism. The symmetry breaking electric field which induces Rashba field is changed in magnitude and direction. Random spatial variation of the Rashba field causes little difference in temporal spin relaxation characteristics in the regime of high-field transport. Spin does not relax monotonically with position but a non-monotonic chaotic trend is observed when Rashba field is varied randomly in magnitude or direction. Chapter 3 discusses electrochemical self-assembly technique employed to fabricate highly ordered nanowires. This nanofabrication technique can be used to grow highly ordered and size controlled nanowires and quantum dots of different materials. Cadmium sulfide nanowires are grown using this method and their infrared photodetectivity is studied. We observe twice as much increase in current under illumination at any given bias.

The study was conducted to evaluate and compare performance of unimodal and bimodal mats with approximately same mass. 10% and 18% Nylon 4, 6 polymer solution were used for electrospinning the fibers. A negative ion source was used to neutralize the surface charge. The fiber diameters were measured with SEM and were <500 nm thus incorporating the slip effect. Bimodal mats were prepared from different deposition modes. Optimal mode was selected on analyzing the performance factors. The bimodal mats were then compared with unimodal mats. For their performance the fiber mass for these mats was approximately the same. It was observed that the unimodal mats had higher efficiencies and higher pressure drop giving a lower FOM. Bimodal mats showed lower efficiencies and pressure drop compared to unimodal mats. However, the FOM for bimodal mats was approximately 200% higher than unimodal mats

Micro-forming studies have been more attractive in recent years because of miniaturization trend. One of the promising metal forming processes, micro-stamping, provides durability, strength, surface finish, and low cost for metal products. Hence, it is considered a prominent method for fabricating bipolar plates (BPP) with micro-channel arrays on large metallic surfaces to be used in Proton Exchange Membrane Fuel Cells (PEMFC). Major concerns in micro-stamping of high volume BPPs are surface interactions between micro-stamping dies and blank metal plates, and tribological changes. These concerns play a critical role in determining the surface quality, channel formation, and dimensional precision of bipolar plates. The surface quality of BPP is highly dependent on the micro-stamping die surface, and process conditions due to large ratios of surface area to volume (size effect) that cause an increased level of friction and wear issues at the contact interface. Due to the high volume and fast production rates, BPP surface characteristics such as surface roughness, hardness, and stiffness may change because of repeated interactions between tool (micro-forming die) and workpiece (sheet blank of interest). Since the surface characteristics of BPPs have a strong effect on corrosion and contact resistance of bipolar plates, and consequently overall fuel cell performance, evolution of surface characteristics at the tool and workpiece should be monitored, controlled, and kept in acceptable ranges throughout the long production cycles to maintain the surface quality. Compared to macro-forming operations, tribological changes in micro-forming process are bigger challenges due to their dominance and criticality. Therefore, tribological size effect should be considered for better understanding of tribological changes in micro-scale. The integrity of process simulation to the experiments, on the other hand, is essential. This study describes an approach that aims to investigate the surface topography changes during long-run micro-stamping of BPPs, and establish relationships between surface roughness–corrosion resistance and surface roughness–contact resistance characteristics of BPPs. Formability levels of formed BPPs and repeatability characteristics of the process were investigated. In addition, blank thickness changes, von-Mises stress, plastic strain levels and distributions of micro-stamping process were determined via finite element analysis (FEA). Test results revealed that the surface roughness change for the stamping dies and BPPs was unsteady (no trend) due to the continuous change of surface topography (i.e. asperity deformation). Sub-micron range local plastic deformations on stamping dies led to surface topography changes on BPP in long-run manufacturing case. As surface defects trigger corrosion, the correlation between surface roughness and corrosion resistance of BPPs was found to be direct. Increasing number of surface irregularities (asperities) lowered contact surface area that resulted in increased contact resistance. ZrN coated BPPs, on the other hand, did not change surface roughness, however; it improved the protection of BPPs against corrosion significantly. In addition, ZrN coating increased the conductivity of BPPs and reduced the contact resistance between BPP and gas diffusion layer (GDL), at certain extent. As dimensional stability and repeatability was confirmed in forming of both uncoated and coated BPPs during the long run manufacturing, different formability levels were achieved for coated and uncoated samples. Lower channel height values were obtained for coated plates because of the different surface hardness of uncoated and coated plates. In tribological size effect part of study, micro stamping experiments using three different dies with distinct channel height values at different stamping force levels were performed. It was concluded that decrease in forming die dimensions led to increase in coefficient of friction as previously reported by other researchers as one of the consequences of tribological size effect. On the other hand, coefficient of friction values were not affected by the force levels used in the experiments and simulations, whereas plastic strain, equivalent stress, and formability levels were increased with increasing stamping force, as expected. In essence, this study proposed a methodology to investigate the long-run manufacturing effects on dimensional stability and surface characteristics of micro-stamped sheets. It also correlates these parameters to fuel cell performance measures such as interfacial contact and corrosion resistance.

Mechanical circulatory support options for patients with a failing Fontan physiology are significantly limited. This research advances the design of a cavopulmonary assist device as a bridge-to-transplant or bridge-to-recovery for Fontans. A fixed-bladed impeller and diffuser for this pump was designed and optimized using ANSYS CFX™ 12.1 software. Building upon the fixed bladed geometry, a novel flexible impeller prototype was created and evaluated by hydraulic testing. In contrast to the fixed design, the flexible impeller enabled a range of blade angles from 60°-150°. Improvement in pump performance was achieved. Pressure generation was found to decrease as a function of higher flow rates and increase as a function of faster rotational speeds and larger blade angles. The designs were able to produce 1-25 mmHg for 0.5-4 L/min at 5000-8000 RPM, which is sufficient to assist Fontan patients. The findings support the continued development of this blood pump with pitch-adjusting characteristics.

Challenges in the development of successful cell therapies involve engineering and control of cues to regulate the balance between differentiation and self-renewal. However, the complexity of architecture and function make this an intriguing problem in the context of forming functional connections. Here we present the design and fabrication of microstructured scaffolds that present a biomimetic framework along which neural cell lines can organize into oriented constructs. Specifically, we show microfabricated non-linear architectures that promote cellular fate related to propagation of human neuroblastoma cells and directed differentiation towards neurons. By mimicking biological networks that allow for spreading of the cells instead of confining them in a groove or a well, a nonlinear configuration can promote a relaxed, self-supportive cell niche. The tailoring of non-homogeneous adhesion sites via the geometry and the compliance and roughness of the substrate allows a versatile microenvironment that promotes propagation and neuronal differentiation.

Today’s real-time systems need to be faster and more powerful than ever before. Caches are an architectural feature that helps solve this problem. Caches however are unpredictable and do not improve the worst case execution time. This work studies the effects of cache-locking on performance and time predictability. Two locking methods were evaluated: a dynamic locking method and a static locking method. The performance of single and multi-core processors and multiple levels of caches were studied. The time predictability of the single core system was studied and the cost of the time predictability was determined for each locking method. Cache-locking in the Level 2 cache had the best performance and the static-locking method had the highest predictability.

Engineering the next generation of smart materials will require new methods of surface characterization, analysis and identification that can be performed not only in three dimensional space but also in the temporal dimension. Of particular interest is the understanding of mechanical properties of complex systems at the micro and nanoscales. Current techniques for such measurements are hampered by challenges including their inability to probe systems in complex microenvironments, non-destructively, or at nanometer resolution. This thesis outlines work in the development of techniques to study diverse systems and determine their mechanical properties using the Atomic Force Microscope as the primary tool. We develop a strategy wherein topography and nanomechanical properties can be simultaneously mapped out to obtain a 3D visualization of a surface at sub micrometer resolutions. A diverse set of applications ranging from polymeric surfaces to protein assembly are studied using this method. In addition to uncovering fundamental surface properties, the groundwork for applying nanomechanical identification for new applications such as forensic identification for bacterial spores are also laid out using this versatile technique.