Computational Nanofluid Flow and Heat Transfer Analyses Applied to Micro-systems

by Koo, Junemo

The compactness and high surface-to-volume ratios of microscale liquid flow devices make them attractive alternatives to conventional flow systems for heat transfer augumentation, chemical reactor or combustor miniaturization, aerospace technology implementations, as well as biomedical applications, such as drug delivery, DNA sequencing, and bio-MEMS, to name a few. While experimental evidence indicates that fluid flow in microchannels, especially in terms of wall friction and heat transfer performance, differs from macrochannel flow behavior, laboratory observations are often inconsistent and contradictory. Some researchers attributed the deviations to unknown microscale effects, which often turned out to originate from inappropriate approaches to analyze the new phenomena. Specifically, system parameters were neglected, which are not important on the macroscale but play important roles in microscale analyses.

The main objectives of the study are to identify important parameters for microscale liquid flows and nanoparticle suspensions, to find a physically sound way to analyze the new phenomena, and to provide mathematical models to simulate them.

Scale analysis was found to be a valuable tool to determine which forces become important on the microscale. With increasing system miniaturization surface forces, such as surface tension and van der Waals forces, take over the control from body forces like gravity and pressure. Furthermore, surface roughness, viscous dissipation, and entrance region effects are very important liquid flow parameters in microscale conduits. In summary, for liquid flow in microchannels with a characteristic width or height of L ¡Ã 10 [µm], the continuum approach, in conjunction with appropriate closure models, is appropriate to analyze microscale effects.

Employing the porous medium layer (PML) idea, surface roughness effects on momentumand heat-transfer in micro-conduits were numerically investigated and verified with experimental data. The friction factor and Nusselt number either increase or decrease depending on the PML model parameters, expressed in terms of the relative surface roughness, Darcy number, Reynolds number, and effective thermal conductivity. Variations in the viscous dissipation effect were found to increasingly affect the friction factor and Nusselt number with decreasing system size. Variations in entrance geometry may cause early laminar-toturbulent transition resulting in higher friction factor values. When nanoparticles are added to liquid flow systems, scalar transport properties can be significantly enhanced. Specifically, nanofluids, i.e., dilute suspensions of nanoparticles in liquids, are used to enhance heat transfer performance or to maximize drug delivery. Focusing on micro-scale heat transfer, it was found that the particle Brownian motion and the induced surrounding liquid motion are key mechanisms for the experimentally observed high increase of the effective thermal conductivity of nanofluids. A new, experimentally validated effective thermal conductivity model has been developed based on kinetic theory. The model predicts both the effective thermal conductivity and dynamic viscosity of nanofluids in terms of nanoparticle concentration, size, density and their interaction potential as well as the density, thermal capacity and dielectric constant of the base liquid. Nanofluid flow applications were tested for micro heat-exchangers and a drug delivery system. Concerning micro heat-exchangers, it was found that a base fluid of high Prandtl number together with nanoparticles of high thermal conductivity in a channel of high aspect ratio, form a desirable combination for optimal performance. In order to minimize the problem of non-uniform suspensions, the selection of materials for the carrier fluid, nanoparticles and conduit wall was found to be very important. For example, the dielectric constants difference should be kept small, and appropriate surface treatment, by creating either electrostatic or steric forces to maintain enough repulsion potential, should be provided.

In the case of nanotherapeutics, radial diffusion turns out to be the controlling mechanism for drug delivery. Specifically, the time scale for radial diffusion should be kept small enough to ensure efficient delivery. A shallow channel design together with a pressure chamber to switch between drug delivery- and purging-process, is suggested over the alternative multi-stream design or the fluid-guiding system proposed by others.