In the recent years, there has been a growing interest in micro- and nano-structured composite systems due to their wide use in microelectronics, optoelectronics, magneto-optical devices, high-density data storage, sensors, biomedical devices, and many other areas. Of particular interest is application in the integrated circuit (IC) industry. Here the need for miniaturization has led to new architectures that combine disparate technologies. This has been achieved through innovations in packaging technologies such as 3D integration for high interconnection density, low power, high data throughput, good signal integrity and reliability, and low cost. One of the key active manufacturing technologies for 3D integration is through silicon vias (TSVs), which involves etching of deep vias in a silicon substrate that are filled with an electrodeposited metal, and subsequent removal of excess metal by chemical mechanical planarization (CMP). Electrodeposition often results in undesired voids in the TSV metal fill as well as a thick overburden layer. These via plating defects can severely degrade interconnect properties and lead to variation in via resistance, electrically open vias, and trapped plating chemicals that present a reliability hazard. Thick overburden layers result in lengthy and expensive CMP processing. We are proposing a technique that pursues a viable method of depositing a high quality metal inside vias with true bottom-up filling, using an additive-free deposition solution. The mechanism is based on a novel concept of electrochemical oxidation of backside silicon that releases electrons, and subsequent chemical etching of silicon dioxide for regeneration of the surface. Electrons are transported through the bulk silicon to the interface of the via bottom and the deposition solution, where the metal ions accept these electrons and electrodeposit resulting in the bottom-up filling of the large aspect ratio vias. With regions outside the vias covered bydielectric, no metal electrodeposition should occur in these regions, which minimizes the metal CMP step and reduces the overall processing times and costs. Hence, inherent bottom-up filling is financially advantageous because it will eliminate a large portion of the metal overburden and associated planarization costs. Additive-free deposition is preferable from both lower production cost and quality management perspectives since it results in higher reliability of deposited metal. Our new bottom-up technique was initially examined and successfully demonstrated on blanket silicon wafers and shown to supply electrons to provide bottom-up filling advantage of through-hole plating and the depth tailorability of blind vias. In order to understand the driving mechanism and limits of this process, we have also conducted a fundamental study that investigated the effect of various process parameters on the characteristics of deposited Cu and Ni and established correlations between metal filling properties and various electrochemical and solution variables. A copper sulfate solution with temperature of about 65 °C was shown to be suitable for achieving stable and high values of current density that translated to copper deposition rates of ~2.4 μm/min with good deposition uniformity. The importance of backside silicon oxidation and subsequent oxide etching on the kinetics of metal deposition on front side silicon has also been highlighted. Further, a process model was also developed to simulate the through silicon via copper filling process using conventional and contactless electrodeposition methods with no additives being used in the electrolyte solution. A series of electrochemical measurements were employed and integrated in the development of the comprehensive process simulator. The experimental data not only provided the necessary parameters for the model but also validated the simulation accuracy. From the simulation results, the “pinch-off” effect was observed for the additive-free conventional deposition process, which further causes partial filling and void formation. By contrast, a void-free filling with higher deposition rates was achieved by the use of the contactless technique. Moreover, experimental results of contactless electrodeposition on patterned wafers showed fast rate bottom-up filling (~3.3 μm/min) in vias of 4 μm diameter and 50 μm depth (aspect ratio = 12.5) without void formation and no copper overburden in the regions outside the vias. Efforts were also made to extend the use of the contactless technique to other applications such as synthesis of porous silicon, which is known to be an excellent material with fascinating physical and chemical properties. We were able to fabricate porous silicon with a morphological gradient using a novel design of the experimental cell. The resulted porous silicon layers show a large distribution in porosity, pore size and depth along the radius of the samples. Symmetrical arrangements were attributed to decreasing current density radially inward on the silicon surface exposed to surfactant containing HF based etchant solution. The formation mechanism as well as morphological properties and their dependence on different process parameters, such as HF concentration, solution pH, surfactant concentration, current density and wafer resistivity, has been investigated in detail. In the presence of surfactants, an increase in the distribution range of porosity, pore diameter and depth was observed by increasing HF concentration or lowering pH of the etchant solution, as the formation of pores was considered to be limited by the etch rates of silicon dioxide. Gradient porous silicon was also found to be successfully formulated both at high and low current densities. Interestingly, the morphological gradient was not developed when dimethyl sulfoxide (instead of surfactants) was used in etchant solution potentially due to limitations in the availability of oxidizing species at the silicon-etchant solution interface. In the last part of the dissertation, we have discussed the gradient bottom up filling of Cu in porous silicon substrates using the contactless electrochemical method. The radially symmetric current that gradually varied across the radius of the sample area was achieved by utilizing the modified cell design, which resulted in gradient filling in the vias. Effect of different deposition parameters such as applied current density, copper sulfate concentration and etching to deposition area ratio has been examined and discussed. Increasing the current density from 10 to 15 mA/cm2 resulted in bottom up deposition with less sharp gradients. Further, the study on the effect of copper sulfate concentration highlighted the importance of mass transfer in this process, as either bottom-up deposition or gradient filling could not be achieved at lower CuSO4 concentrations (0.1 and 0.25 M). Additionally, the filling gradient of deposited Cu was obtained with etching to deposition area ratio of 1.6 and 2.7, while a more uniform deposition was observed when the ratio was increased to 3.8. This suggested that the gradient filling may only be accomplished within a certain range of the etching to deposition area ratios.

This experimental research aims at investigating the single-phase flow heat transfer and friction factor, flow boiling heat transfer and pressure drop, and flow visualisation in microchannels using de-ionized water. In the literature, many studies failed to explain the effect of aspect ratio on the single-phase and two-phase flow heat transfer rate and pressure drop. Because the channel aspect ratios and hydraulic diameters were varied together in those studies. Also, there is a discrepancy between past studies and the conventional theory for the flow boiling heat transfer characteristics. Accordingly, the objectives of this research can be listed as follows: (i) modifying the existing experimental facility to perform single-phase and two-phase flow heat transfer and pressure drop and two-phase flow pattern visualization experiments in microchannels, (ii) clarifying the fundamental aspects of flow boiling in micro passages, (iii) investigating the aspect ratio, heat flux, mass flux and vapour quality effects on flow patterns, heat transfer rate and pressure drop in single-phase and two-phase flow, (iv) comparing the obtained results with heat transfer and pressure drop correlations and flow pattern maps available in the literature. Consequently, the pre-existing experimental facility was modified in the current research by changing the pre-heaters, flowmeter and piping in order to achieve the goals of this study. Four copper rectangular microchannels were designed and manufactured. Three microchannel test sections having the same hydraulic diameter and length but different aspect ratios were investigated to reveal the effect of aspect ratio on the single-phase and two-phase flow heat transfer rate and pressure drop. The surface roughness of each microchannel was also examined. It was found that the surface roughnesses of all microchannels are similar. Moreover, an additional microchannel test section was used to examine the effect of heated length on the flow boiling heat transfer coefficient and pressure drop. The single-phase flow results demonstrated that the channel aspect ratio has no influence on the friction factor and heat transfer rate for the tested microchannels and experimental range. In the flow boiling experiments, bubbly, bubbly/slug, slug, churn and annular flow regimes were observed in the tested microchannels. The channel aspect ratio effect was found to be small on the observed flow patterns. The experimental flow patterns were predicted well by the flow pattern map proposed by Galvis and Culham (2012) except for the slug flow regime. The flow pattern maps of Sobierska et al. (2006) and Harirchian and Garimella (2009) reasonably predicted the experimental flow pattern data. The flow boiling heat transfer results showed that the prevailing heat transfer mechanism is nucleate boiling for the low and medium heat flux inputs. On the other hand, the dominant heat transfer mechanism is unclear at the high heat flux inputs while smaller aspect ratio microchannel has better heat transfer performance for low and medium heat flux inputs. However, at high heat flux inputs the channel aspect ratio effect was found to be insignificant on the flow boiling heat transfer coefficient. The experimental flow boiling heat transfer coefficient data were reasonably predicted by the correlations of Sun and Mishima (2009), Li and Wu (2010) and Mahmoud and Karayiannis (2011) from the literature. The flow boiling pressure drop characteristics were also examined in the tested microchannels. Outcome of the experiments consistently indicated a highly linear trend between the increasing flow boiling pressure drop and the heat and mass flux. Also, the flow boiling pressure drop increased with the increase in vapour quality. The effect of channel aspect ratio on the flow boiling pressure drop was also assessed. It was found that when the channel aspect ratio decreased, the flow boiling pressure drop increased. The experimental flow boiling pressure drop data were compared to correlations from the literature. Mishima and Hibiki (1996), Yu et al. (2002) and Zhang et al. (2010) correlations reasonably predicted the experimental flow boiling pressure drop results.

Recent tsunami disasters caused devastating damages to well-engineered coastal infrastructures. In fact, the current design guidelines are not able to provide realistic estimations of tsunami loads in order to design structures to withstand tsunamis. Tsunami hydrodynamic forces are estimated using the drag coefficient. This coefficient is traditionally calculated based on a steady flow analogy. However, tsunami bores behave like unsteady flows. The present work aims at investigating the tsunami forces for different structure geometries to provide realistic guidelines to estimate drag coefficients considering unsteady flows. In the present paper, the dam-break approach is used to investigate the tsunami-like bore interaction with structures. A three-dimensional multiphase numerical model is implemented to study the tsunami induced forces on rectangular shape structures with various aspect ratios (width/depth) and orientations. The numerical model results are validated using measured forces and bore surface elevations of the physical experiments. A scaled-up domain is modeled in order to eliminate the effects of domain sidewalls in the simulation results. The drag coefficient relations with structure geometries and bore depths are provided. The obtained hydrodynamic forces and drag coefficients are compared with existing data in the literature and design codes. For the second topic, a multi-phase three-dimensional numerical reproduction of a large scale laboratory experiment of tsunami-like bores interaction with a surface-piercing circular column is presented. The numerical simulation is conducted in OpenFOAM. The dam-break mechanism is implemented in order to generate tsunami-like bores. The numerical model is validated using the experimental results performed at Canadian Hydraulics Center of the National Research Council (NRC-CHC) in Ottawa. The unsteady Reynolds Averaged Navier-Stokes equations (RANS) are used in order to treat the turbulence effects. The Shear Stress Transport (SST) k-ω turbulence model showed high level of accuracy in replication of the bore-structure interaction. Further, a scaled-up domain is used to investigate the influence of the bed condition in terms of various downstream depths and roughness. Finally, a broad investigation on the bore propagation characteristics is performed. The resulting stream-wise forces exerted on the structural column as well as the bore velocity are compared and analyzed for smooth, rough, dry and wet beds with varying depths.

No description available.

Recent technological developments, mostly in the fields of concentrated solar power and microelectronics, have driven heat transfer requirements higher than current heat exchangers are capable of producing. Processing power is increasing, while processor size simultaneously decreases and the heat flux requirements of concentrating solar power plants are being driven up by the high temperatures that produce the best thermal efficiency. Heat transfer in microchannels, specifically when utilising flow boiling, has been shown to produce significantly higher heat fluxes than their macro-scale counterparts and could have a large impact on many industrial fields. This high heat transfer characteristic is caused by a number of factors, including the large difference between the sensible and latent heat of the working fluid and the evaporation of a thin liquid film that forms between the microchannel walls and the vapour bubbles. These phenomena occur at incredibly small scales. Flow visualisations, temperature and pressure measurements are therefore difficult to obtain. Many experiments that cover a wide range of microchannel sizes, shapes and orientations, and utilise different working fluids and heat fluxes have been reported. However, the correlations between confined boiling, heat flux and pressure drop have mostly been produced for macro-scale flow. Many different criteria have been developed to distinguish the macro scale from the micro scale, but the general consensus is that macro-scale heat transfer correlations do not perform well when used in the micro scale. Heat transfer correlations are typically created by performing physical experiments over a wide range of parameters and then quantifying the effect that varying these parameters has on the performance of the system. The small scale and high complexity of microchannel-based heat exchangers make visualising the flow within them difficult and inaccurate because both the working fluid and the microchannel walls distort light. The use of numerical modelling via computational fluid dynamics software allows phenomena that occur within the channel to be simulated, which provides valuable insight into how rapid bubble growth affects the surrounding fluid, which can lead to the design of better heat exchangers. This study focused on numerically modelling the growth of a single bubble during the flow boiling of FC-72 in a microchannel with a hydraulic diameter of 0.9 mm and an aspect ratio of 10. The numerical domain was limited to a 10 mm section of the microchannel where bubble nucleation and detachment were observed in an experimental study on a similar microchannel setup. The high cost of 3D simulations was offset by an interface-tracking mesh refinement method, which refined cells not only at the interface, but also a set distance on either side of the interface. To focus on the effects of gravity, a simplified approach is used, which isolates certain phenomena. Density gradients, material roughness and multiple bubble interaction are ignored so that the effects of buoyancy and bubble detachment can be analysed. Simulations are first performed in a 2D section through the centre of the microchannel, and then in the full 3D domain. In both the 3D numerical and experimental cases (Meyer et al., 2020), the bottom heated case had the lowest maximum temperature and the highest heat transfer characteristics, which were influenced by the detachment of the bubble from the heated surface. This observation indicates that the gravitational orientation of the channel can have a significant effect on the heat transfer characteristics of microchannel-based heat exchangers, and that more investigation is required to characterise the extent of this effect. / Dissertation (MEng)--University of Pretoria, 2019. / Mechanical and Aeronautical Engineering / MEng / Unrestricted

No description available.

Bats are important components of biodiversity within forested ecosystems. This research addressed habitat characteristics that influence species occupancy and stable isotopes and wing morphology to assess community structure within the Sam D. Hamilton Noxubee National Wildlife Refuge. To meet these objectives, I deployed echolocation recorders, mist-nets and conducted roost checks to capture bat acoustics; fur samples were also collected to measure ratios of carbon (C13/12) and nitrogen (N15/14). Relationships between occupancy, habitat class and features were not apparent for most species. However, Lasiurus and Mytois spp. showed positive relationships with proximity to roads, Lasiurus, positive with stem density and Perimyotis subflavus, negative with basal area. Stable isotope analysis revealed some distinction of trophic niches while wing morphometrics indicated bats of similar wing shape and size show greater trophic overlap. Collectively, these results suggest that habitat management, as current within the study area, will have limited influence on local bat distributions.

This study demonstrates the applicability of using a modified application strategy of panel method to analyze low aspect ratio wings at preliminary design phases. Conventional panel methods fail to capture the leading edge vortex (LEV) that is shed by wings with low aspect ratios, typically below 2 depending on planform. This aerodynamic phenomenon contributes to a significant amount of the lift of these wings and the result is a drastic underestimation of the lift characteristics when analyzed by conventional panel method. To capture the effect of the leading edge vortex, a panel method code was used with an extended definition of the Kutta condition along portions of the leading edge inducing a vortex to shed from the leading edge and flow aft just inside the leading edge. To validate that this method, it was applied to 2 elliptical planforms with constant thickness where experimental force balance data was available. Additionally, the same 2 wings were analyzed using a finite volume solver to compare pressure distributions and to demonstrate the difference in magnitude of solution times. For comparison purposes, the resulting forces and moments from both computational methods and experimental testing were plotted over a range of angles of attack. Overall, the results demonstrate that a modified panel method could be used during the preliminary design phases for low aspect ratio wings. The panel method can reasonably model the lift and induced drag characteristics of low aspect ratio wings. This method loses applicability beyond the stall point where the leading edge vortex breaks down and oversimplifies pitching moment relation to angle of attack. Additionally, when compared to finite volume solutions of the same scenario, the panel method provided a result 20 to 30 times faster than the finite volume solutions. With this in mind, the modified panel method application strategy lends itself to preliminary design phases of low aspect ratio wings where the level of detail does not warrant finite volume analysis and solution speed has higher priority.

No description available.

No description available.