Over the past 15 years, research in the field of microfluidics has rapidly gained popularity. By seeking to miniaturize and automate separation-based analysis, microfluidic research seeks to improve current methods through decreased cost, analysis time, and sources of contamination. My work has focused on developing a novel fabrication method, based on standard microfabrication techniques, to create thin-film microfluidic devices. This microfabrication format makes it possible to generate devices that provide high efficiencies, enable mass fabrication, and provide a platform capable of integrating the microfluidic and electronic components necessary for a micro-total analysis system (μ-TAS). Device fabrication combines the processes of photolithography, thermal evaporation, plasma enhanced chemical vapor deposition (PECVD), and wet chemical etching to ultimately provide hollow-core channels. When these microcapillaries are filled with buffer and potentials are applied across them, control of the flow in the channels can be established. By designing intersecting microchannels having an offset “T†geometry, I have been able to inject and electrophoretically separate three fluorescently labeled amino acids and obtain efficiencies of over 2500 theoretical plates. Through the addition of commercially available electroosmotic flow reducing coatings, I have been able to improve the separation of these amino acids, decreasing the run time by approximately 6 fold and increasing the efficiency by as much as 10 fold. Through the use of these coatings I have also been able to carry out electrophoretic separations of three peptides. My most recent work has focused on the polymerization of acrylamide gels in these channels. A method for the selective placement of a gel has been developed using a prepolymer solution with a light-sensitive initiator. Further work to adjust the polymer pore size and interface with ampholyte-containing gels should allow methods such as capillary gel electrophoresis (CGE), preconcentration, and two dimensional (isolectric focusing and CGE) separations to be performed. The development of gel-based analysis methods, along with other fluidic and electrical capacities, should move thin-film microdevices toward the realization of the lab-on-a-chip concept.

Due to the increase of application in a number of emerging technologies, a growing amount of research has focused on the reduction of drag in microfluidic transport. A novel approach reported in the recent literature is to fabricate micro-ribs and cavities in the channel wall that are then treated with a hydrophobic coating. Such surfaces have been termed super- or ultrahydrophobic and the contact area between the flowing liquid and the solid wall is greatly reduced. Further, due to the scale of the micropatterned structures, the liquid is unable to wet the cavity and a liquid meniscus is formed between ribs. This creates a liquid-vapor interface at the cavity regions and renders surfaces with alternating regions of no-slip and of reduced shear on the microscale. This thesis reports the numerical study of hydrodynamically fully-developed laminar and turbulent flows through a parallel plate channel with walls exhibiting micro-ribs and cavities oriented parallel to the flow direction, where fully developed turbulent flow is considered in a time-averaged sense. Three laminar flow models are implemented to investigate the liquid-vapor interface and to account for the effects of the vapor motion in the cavity regions. For each of the laminar flow models, the liquid-vapor interface was idealized as a flat interface. As a benchmark for the proceeding laminar flow models, the first model considers the case of a vanishing shear stress at the interface between the liquid and vapor domains. Effects of the vapor motion in the cavity are then accounted for in a one-dimensional cavity model where the vapor velocity is considered to be dependent on the wall normal coordinate only, followed by a two-dimensional cavity model that accounts for the vapor velocity's dependence on the transverse coordinate as well. The vapor cavity is modeled analytically and is coupled to the liquid domain by equating the fluid velocities and shear stresses at the liquid-vapor interface. In the turbulent flow model the liquid-vapor interface is idealized as a flat interface with a zero shear stress boundary condition. In general the numerical predictions show a reduction in the total frictional resistance as the cavity width is increased relative to the channel width, the channel height-to-width aspect ratio is decreased, and the vapor cavity depth is increased. The frictional resistance is also reduced with increased Reynolds number in the turbulent flow case. In the range of parameters examined for each fluid flow regime, reductions in drag as high as 91% and 90% are reported for the laminar flow and turbulent flow models, respectively. Under similar conditions however, the turbulent flow results indicate a greater reduction in flow resistance than for the laminar flow scenario. Based on an analysis of the obtained data, analytical expressions are proposed for both laminar and turbulent flow which facilitates the prediction of the frictional resistance.

Quantification and prediction of capillary fluid flow in planar nanochannels is essential to the development of many emerging nanofluidic technologies. Planar nanochannels are typically produced using the standard nanofabrication processes of thermal bonding or sacrificial etching. Both approaches may yield nanochannels that are bowed and/or exhibit non-uniform (i.e. non-planar) wall spacing. These variations in wall spacing affect the transient dynamics of a liquid plug filling the nanochannel, causing deviations from the classical behavior in a parallel-plate channel as described by the Washburn model. Non uniform wall spacing impacts the overall frictional resistance and influences the meniscus curvature. In this thesis, a new analytical model that predicts the meniscus location over time in micro- and nanochannels as a function of channel height was compared to experimental filling data of well-characterized channels with different heights. The wall-to-wall spacing of the utilized nanochannels exhibited height variations between 60 and 300 nm. The model was also validated with microscale channels that were fabricated with a linear variation in the wall-to-wall spacing from 100 µm to 400 µm. The filling speed and meniscus shape during the filling process were determined by dynamic imaging of the meniscus front for several different liquids. A modified Washburn equation that utilizes an effective channel height to predict the filling speed corresponding to the location of the tallest height within a channel was derived. A model was also developed to predict the meniscus distortion encountered in a non-constant height channel, provided the cross-sectional channel heights and the distance from the channel entrance are known. The models developed herein account for induced transverse pressure gradients created by non-constant channel heights. The models are compared to experimental data derived from both nanoscale and microscale channels with good qualitative agreement. These results demonstrate that the capillary flow in nanochannels with non-parallel-plate, linear tapered, or parabolic cross sections can be predicted.

Mini-/microchannel components have received attention over the past few decades owing to their compactness and superior thermal performance. Microchannel heat sinks are typically manufactured through traditional manufacturing practices (milling and sawing, electrodischarge machining, and water jet cutting) by changing their components to work in microscale environments or microfabrication techniques (etching and lost wax molding), which have emerged from the semiconductor industry. An extrusion process is used to produce multiport minichannel-based heat exchangers (HXs). However, geometric manufacturing limitations can be considered as drawbacks for all of these techniques. For example, a complex out-of-plane geometry is extremely difficult to fabricate, if not impossible. Such imposed design constraints can be eliminated using additive manufacturing (AM), generally known as three-dimensional (3D) printing. AM is a new and growing technique that has received attention in recent years. The inherent design freedom that it provides to the designer can result in sophisticated geometries that are impossible to produce by traditional technologies and all for the redesign and optimization of existing models. The work presented in this thesis aims to investigate the thermal performance of enhanced minichannel HXs manufactured via metal 3D printing both numerically and experimentally. Rectangular winglet vortex generators (VGs) have been chosen as the thermal enhancement method embedded inside the flat tube. COMSOL Multiphysics, a commercial software package using a finite element method (FEM), has been used as a numerical tool. The influence of the geometric VG parameters on the heat transfer and flow friction characteristics was studied by solving a 3D conjugate heat transfer and laminar flow. The ranges of studied parameters utilized in simulation section were obtained from our previous interaction with various AM technologies including direct metal laser sintering (DMLS) and electron-beam melting (EBM). For the simulation setup, distilled water was chosen as the working fluid with temperaturedependent thermal properties. The minichannel HX was assumed to be made of AlSi10Mg with a hydraulic diameter of 2.86 mm. The minichannel was heated by a constant heat flux of 5 Wcm−2 , and the Reynolds number was varied from 230 to 950. A sensitivity analysis showed that the angle of attack, VG height, VG length, and longitudinal pitch have notable effects on the heat transfer and flow friction characteristics. In contrast, the VG thickness and the distance from the sidewalls do not have a significant influence on the HX performance over the studied range. On the basis of the simulation results, four different prototypes including a smooth channel as a reference were manufactured with AlSi10Mg via DMLS technology owing to the better surface roughness and greater design uniformity. A test rig was developed to test the prototypes. Owing to the experimental facility and working fluid (distilled water), the experiment was categorized as either a simultaneously developing flow or a hydrodynamically developed but thermally developing flow. The Reynolds number ranged from 175 to 1370, and the HX was tested with two different heat fluxes of 1.5 kWm−2 and 3 kWm−2 . The experimental results for the smooth channel were compared to widely accepted correlations in the literature. It was found that 79% of the experimental data were within a range of ±10% of the values from existing correlations developed for the thermal entry length. However, a formula developed for the simultaneously developing flow overpredicted the Nusselt number. Furthermore, the results for the enhanced channels showed that embedding VGs can considerably boost the thermal performance up to three times within the parameters of the printed parts. Finally, the thermal performance of the 3D-printed channel showed that AM is a promising solution for the development of minichannel HXs. The generation of 3D vortices caused by the presence of VGs ii can notably boost the thermal performance, thereby reducing the HX size for a given heat duty.

MEMS (Micro Electro Mechanical Systems) have received a great deal of attention in both the research and industrial sectors in recent decades. The broad MEMS category, microfluidics, the study of fluid flow through channels measured on the micrometer scale, plays an important role in devices such as compact heat exchangers, chemical and biological sensors, and lab-on-a-chip devices. Most of the research has been focused on how entire systems operate, both experimentally and through simulation. This paper strives, systematically, to map them through experimentation of the previous to untested realm of pressure loss through laminar square-profile sweeping-bend microchannels. Channels were fabricated in silicone and designed so a transducer could detect static pressure across a very specific length of channel with a desired bend. A wide variety of Reynolds numbers, bend radii, and bend angles were repeatedly tested over long periods in order to acquire a complete picture of pressure loss with in the domain of experimentation. Nearly all situations tested were adequately captured with the exception of some very low loss points that were too small to detect accurately. The bends were found to match laminar straight-duct theory at Reynolds numbers below 30. As Reynolds numbers increased, however, minor losses began to build and the total pressure loss across the bend rose above straight-duct predictions. A new loss coefficient equation was produced that properly predicted pressure losses for sweeping-bends at higher Reynolds numbers; while lower flow ranges are left to laminar flow loss for prediction.

Heat transfer and flow characteristics of phase change material slurry flow in microchannels with constant heat flux at the base were investigated. The phase change process was included in the energy equation using the effective specific heat method. A parametric study was conducted numerically by varying the base fluid type, particle concentration, particle size, channel dimensions, inlet temperature, base heat flux and melting range of PCM. The particle distribution inside the microchannels was simulated using the diffusive flux model and its effect on the overall thermal performance of microchannels was investigated. Experimental investigation was conducted in microchannels of 101 [micro]m width and 533 [micro]m height with water as base fluid and n-Octadecane as PCM to validate the key conclusions of the numerical model. Since the flow is not fully developed in case of microchannels (specifically manifold microchannels, which are the key focus of the present study), thermal performance is not as obtained in conventional channels where the length of the channel is large (compared to length of microchannels). It was found that the thermal conductivity of the base fluid plays an important role in determining the thermal performance of slurry. The effect of particle distribution can be neglected in the numerical model under some cases. The performance of slurry depends on the heat flux, purity of PCM, inlet temperature of the fluid, and base fluid thermal conductivity. Hence, there is an application dependent optimum condition of these parameters that is required to obtain the maximum thermal performance of PCM slurry flows in microchannels.

Many diseases, such as Parkinson's disease and heavy metal poisoning, are associated with impaired or aberrant locomotion. Because the underlying mechanisms are difficult to study in humans, simpler metazoans like Caenorhabditis elegans are commonly employed to model these diseases. C. elegans is especially useful in this respect because its innate electrotactic behaviour allows instantaneous manipulation of its locomotion using mild electric fields in a microfluidic environment, the results of which can be captured on video. However, extraction of locomotory data from these videos is a major bottleneck to the throughput of the microfluidic electrotaxis platform. In the present study, we describe the development of novel software to analyze electrotaxis videos in an automated fashion. The software, dubbed the Automated Nematode Tracking System (ANTS), uses efficient, parameterless computer vision techniques to simultaneously track and assess movement characteristics of ambulating animals. In combination with the previously described microfluidic electrotaxis platform, ANTS promises to accelerate research with C. elegans models of locomotory dysfunction. / Thesis / Master of Applied Science (MASc)

No description available.

No description available.

No description available.