More than 1.2 million people worldwide require regular hemodialysis therapy to treat end stage renal failure. In the United States alone, there are 300,000 patients and the National Kidney Foundation predicts that this number will double in the next 10 years. Currently most dialysis patients receive treatment at a dialysis center and need three 4-5 hour treatments each week. While these treatments are useful, more frequent and longer duration dialysis better simulates natural kidney function. Consequently, at-home hemodialysis is expected to provide patients a better quality of life. Current hemodialysis systems are too expensive to support at-home hemodialysis. Cost drivers include the capital costs of the hemodialysis equipment and the raw material costs of expensive hemodialysis membranes. Microchannel hemodialysers have smaller form factors requiring significantly less membrane while enabling reductions in the size and cost of capital equipment. Microchannel devices are typically made by microchannel lamination methods involving the patterning, registration and bonding of thin laminae. Findings in this paper show that membrane utilization is highly dependent on registration accuracy with membrane utilization often dropping below 25%. Efforts here focus on the development of a self-registration method for assembling microchannel hemodialysers capable of supporting registration accuracies below 25 ��m over a 50 mm polycarbonate lamina. Using these methods, registration accuracies below 13 ��m were measured over a 50 mm scale. A mass transfer test article was produced with measured average one dimensional misregistration below 19 ��m with a demonstrated membrane utilization of 44.9% when considering both microchannel and header regions. Mass transfer results suggest that the device performed with a mass transfer area of 90.59 mm��. A design is proposed describing membrane utilization of over 79%. / Graduation date: 2013 / Access restricted to the OSU Community at author's request from 9-17-2012 - 9-17-2013

Microchannel process technology (MPT) components are chemical unit operations which exploit highly-parallel arrays of microchannels to process large fluid volumes for portable and distributed applications. Microchannel heat exchangers (MCHXs) have demonstrated 3 to 5 times higher heat fluxes when compared to conventional heat exchangers resulting in proportionate reductions in size and weight. The most common fabrication approach for producing MPT components is microchannel lamination, or microlamination, in which thin layers of metal or polymer are patterned with microchannel features, registered, and bonded to produce monolithic components. Currently, the most common microlamination architecture involves the photochemical machining and diffusion bonding of metal foils. Prior work has established that the yields in diffusion bonding often drive the costs of MCHXs. Laser keyhole welding has been proposed as an alternative bonding technology providing the potential for faster cycle times, smaller weld widths and layer-to-layer evaluation of hermeticity leading to higher yields. Furthermore, laser weldments have small heat-affected zones providing excellent mechanical strength. In this study, efforts are made to evaluate the feasibility of using laser welding in the microlamination of a high-temperature counter-flow heat exchanger made of a Ni superalloy. Preliminary efforts were focused on the development and validation of weld strength estimation models. These models were then used to narrow down the range of process parameters and a final set of process parameters was determined through the use of a full factorial experiment with weld strength, joining efficiency and weld gap as response variables. The most acceptable parameter set was used to demonstrate the fabrication of a Haynes 214 microchannel array with adequate bond strength and hermeticity and minimal thermal warpage. / Graduation date: 2013

A new adhesive bonding method is introduced for microlamination architectures, for producing low-temperature microchannel arrays in a wide variety of metals. Sheet metal embossing and chemical etching processes have been used to produce sealing bosses and flow features, resulting in approximately 50% fewer laminae over traditional methods. These lamina designs are enabled by reduced bonding pressures required for the new method. An assembly process using adhesive dispense and cure is outlined to produce leak-free devices. Feasible fill ratios were determined to be 1.1 in general and 1.25 around fluid headers, largely due to gaps between faying surfaces caused by surface roughness. Bond strength investigation reveals robustness to surface conditions and a bond strength of 5.5-8.5 MPa using a 3X safety factor. Dimensional characterization reveals a two sigma (95%) post-bonded channel height tolerance under 10% (9.6%) after bonding. Patterning tolerance and surface roughness of the faying laminae were found to have a significant influence on the final postbonded channel height. Leakage and burst pressure testing on several samples has established confidence that adhesive bonding can produce leak-free joints. Operating pressures up to 413 kPa have been satisfied, equating to tensile pressure on bond joints of 1.9 MPa. Higher operating pressures can be accommodated by increasing the bond area of devices. A two-fluid counterflow microchannel heat exchanger has been redesigned, fabricated and tested to demonstrate feasibility of the new method. Results show greater effectiveness and higher heat transfer rates, suggesting a smaller device than the original heat exchanger. A maximum effectiveness of 82.5% was achieved with good agreement between theoretical and experimental values. Although thermal performance was improved, higher pressure drops were noted. Pressure drops were predicted with a maximum error of 16% between theoretical and experimental values. Much of the pressure drop was found to be in the device manifolds, which can be improved in subsequent designs. Fluid flow simulation results show a 45-65X reduction in fluid leakage velocity past sealing bosses, thereby mitigating adhesive erosion concerns. Theoretical models indicate that the worst-case adhesive erosion rate is 1/12th the rate of aluminum and 1/7th the rate of stainless steel, implying satisfactory reliability in high fluid velocity applications. Economic comparison indicates an 83% reduction in material cost and 71% reduction in assembly cost with the new adhesive bonding process, when compared to diffusion bonding for the recuperator investigated in this study. Adhesive compatibility with common refrigerants is reviewed through literature references, with no adverse compatibility issues noted. The findings of this research suggest a fairly quick path to commercialization for the new bonding method. Future studies required to pursue commercialization are liquid and gas permeability evaluations, and long term strength and performance testing of adhesives in targeted applications. / Graduation date: 2012 / Access restricted to the OSU Community at author's request from Mar. 26, 2012 - Mar. 26, 2013

Microchannel process technology (MPT) offers several advantages to the field of nanomanufacturing: 1) improved process control over very short time intervals owing to shorter diffusional distances; and 2) reduced reactor size due to high surface area to volume ratios and enhanced heat and mass transfer. The objective of this thesis was to consider how nanomaterials, produced in part using MPT, could be used to solve problems associated with the fabrication of MPT devices. Specifically, many MPT devices are produced using transient liquid-phase brazing involving an electroplated interlayer consisting of a brazing alloy designed for melting temperature suppression. Unfortunately, these alloys can form brittle secondary phases which significantly reduce bond strength. In contrast, prior efforts have shown that it is possible to leverage the size-dependent properties of nanomaterials to suppress brazing temperatures. In this prior work, thin films of off-the-shelf elemental nanoparticles were used as interlayers yielding joints with improved mechanical properties. In the present investigation, efforts have been made to characterize the synthesis and deposition of various elemental nanoparticle suspensions for use in the transient liquid-phase brazing of aluminum and stainless steel. Advances were used to demonstrate the nanoparticle-assisted diffusion brazing of a microchannel array. In the first section, a silver nanoparticle (AgNP) interlayer was produced for the diffusion brazing of heat exchanger aluminum. Efforts are made to examine the effect of braze filler particle size (~5 nm and ~50 nm) and processing parameters (heating rate: 5ºC/min and 25ºC/min; brazing temperature: 550ºC and 570ºC) on thin coupons of diffusion-brazed 3003 Al. A tensile strength of 69.7 MPa was achieved for a sample brazed at 570°C for 30 min under 1 MPa with an interlayer thickness of approximately 7 μm. Further suppression of the brazing temperature to 500ºC was achieved by sputtering a 1 µm thick layer of Cu before depositing a 5 nm thick film of AgNPs resulting in a lap shear strength of 45.3±0.2 MPa. In the middle section of this thesis, several techniques are investigated for the synthesis of sub 10 nm diameter nickel nanoparticles (NiNPs) to be used in the diffusion brazing of 316L stainless steel. The average NiNP size was varied from 9.2 nm to 3.9 nm based on the synthesis technique, solvent and reducing agent used. Conventional wet-chemical synthesis using NiCl₂.6H₂O in ethylene glycol (solvent) and N₂H₄.H₂O (reducing agent) resulted in the formation of 5.4 ± 0.9 nm NiNPs. Continuous flow synthesis using a microchannel T-mixer (barrel diameter of 521µm) and a 10 second residence time of reactants in a bath temperature of 130ºC resulted in a particle size of with 5.3 ± 1 nm. To make the synthesis safer and less energy intense, microwave heating was used along with less toxic Ni(CH₃CO₂)₂·4H₂O (nickel salt), propylene glycol (solvent) and NaPH₂O₂ (reducing agent) yielding 3.9 ± 0.8 nm diameter NiNPs. For the final section, nickel nanoparticles were synthesized using NiCl₂.6H₂O (nickel salt), de-ionized water (solvent), NaBH₄ (co-reducing agent), N₂H₄.H₂O (reducing agent) and polyvinylpyrolidone (capping agent) yielding 4.2 ± 0.6 nm NiNP. Several deposition techniques were investigated for controlling film thickness and uniformity in the diffusion brazing of 316L stainless steel (SS). Using in-house prepared NiNP and automated dispensing, a hermetic joint up to 70 psi (tested pressure) was obtained in 316L SS substrates under brazing conditions of 800ºC, 2 MPa and 30 min. Throughout the course of this thesis, techniques used for characterizing nanoparticles, films and joints included FT-IR, XRD, SEM, TEM, HRTEM, EDS, EPMA, DSC, mass spectrometry, and lap-shear testing. / Graduation date: 2012

Rising energy demands and the continual push to find more energy efficient technologies have been the impetus for the investigation of waste heat recovery techniques. Diesel engine exhaust heat utilization has the potential to significantly reduce the consumption of fossil fuels and reduce the release of greenhouse gases, because diesel engines are ubiquitous in industry and transportation. The exhaust energy can used to provide refrigeration by implementing an organic Rankine cycle coupled with a vapor-compression cycle. A critical component in this system, and in any waste heat recovery system, is the heat exchanger that extracts the heat from the exhaust. In this study, a cross-flow microchannel heat exchanger was geometrically examined and thermally tested under laboratory conditions. The heat exchanger, referred to as the Heat Recovery Unit (HRU), was designed to transfer diesel exhaust energy to a heat transfer oil. Two methods were developed to measure the geometry of the microchannels. The first was based on image processing of microscope photographs, and the second involved an analysis of profilometer measurements. Both methods revealed that the exhaust channels (air channels) were, on average, smaller in cross-sectional area by 11% when compared to the design. The cross-sectional area of the oil channels were 8% smaller than their design. The hydraulic diameters for both channel geometries were close to their design. Hot air was used to simulate diesel engine exhaust. Thermal testing of the heat exchanger included measurements of heat transfer, effectiveness, air pressure drop, and oil pressure drop. The experimental results for the heat transfer and effectiveness agreed well with the model predictions. However, the measured air pressure drop and oil pressure drop were significantly higher than the model. The discrepancy was attributed to the model's ideal representation of the channel areas. Additionally, since the model did not account for the complex flow path of the oil stream, the measured oil pressure drop was much higher than the predicted pressure drop. The highest duty of the Heat Recovery Unit observed during the experimental tests was 12.3 kW and the highest effectiveness was 97.8%. To examine the flow distribution through the air channels, velocity measurements were collected at the outlet of the Heat Recovery Unit using a hot film anemometer. For unheated air flow, the profile measurements indicated that there was flow maldistribution. A temperature profile was measured and analyzed for a thermally loaded condition. / Graduation date: 2012