• Computational analysis of nanofluids flow and heat transfer in microchannels and fin tube air coils

      Ray, Dustin R.; Das, Debendra K.; Peterson, Rorik A.; Misra, Debasmita; Kim, Sunwoo (2021-05)
      The four goals of this dissertation were to investigate nanofluids' thermal and fluid dynamic performance in (i) an air coil, (ii) microchannel heatsink, using computational fluid dynamic (CFD) software, ANSYS Fluent, develop (iii) hydrodynamic entrance length correlation and (iv) apparent friction factor correlations in rectangular microchannels. In cold regions of the world, ethylene glycol mixed with water (EG/W) are used as a heat transfer fluid instead of water due to their freeze protection. EG/W has low thermal conductivity than water, which can be improved by dispersing nanoparticles and creating a new fluid called nanofluid. A computational scheme was developed based on the Effectiveness-Number of Transfer Unit (ε-NTU) method to compare nanofluids' thermal and fluid dynamic performance to the conventional ethylene glycol and water mixture. The nanofluid's performance was examined by conducting two studies: reducing pumping power and reducing the air coil's surface area via length. The results showed at a dilute concentration of 1% of Al₂O₃ can reduce the pumping power requirements by 35.3% or reduce the air coil length by 7.4% while maintaining the same heat transfer rate as EG/W. The results show nanofluids could provide significant savings in energy or material costs. The nanofluids' (Al₂O₃, CuO, and SiO₂) thermal and fluid dynamic performance used in a microchannel heatsink was explored using analytical and computational methods. The computational model was developed in ANSYS Fluent. Comparing analytical and computational results, good agreement was observed validating both methods. The three nanofluids had a maximum difference of 4.1% for pressure drop and 2.9% for the Nusselt number. Three performance studies were conducted using the analytical model based on constant Reynolds number, maximum surface temperature, and pumping power. A constant Reynolds number of nanofluids could reduce the maximum surface temperature by 6K, but at the cost of increased pumping power. Nanofluids showed the pumping power could be reduced by 23% compared to the base fluid while maintaining equal maximum surface temperature. In electronic cooling applications where microchannel heatsinks are used, nanofluids seem promising for lowering critical components' operating temperatures and contribute to increased life and system reliability. A detailed three-dimensional laminar flow CFD model was developed and ran for Reynolds numbers ranging from 0.1 to 1000 through six rectangular microchannels aspect ratios (α): 1, 0.75, 0.5, 0.25, 0.2, 0.125. The majority of the Reynolds numbers simulated were in the low regime (Re< 100) to fulfill the lack of literature for determining accurate hydrodynamic entrance length and apparent friction factor for microchannels. From these numerical simulations, improved correlations were developed to predict hydrodynamic entrance length with a mean error of less than 2% and a maximum error of 5.75% for 0.1 ≤ Re ≤ 1000 & 0 ≤ α ≤ ∞. For the apparent friction factor in microchannels, three correlations were derived from the numerical simulations: fully developed friction factor (fRe), developing incremental pressure drop number (K(z)), and fully developed incremental pressure drop (K(∞)). The three correlations were used to determine the local fapp,zRe, in the applicable range of 0.1 ≤ Re ≤ 1000 & 0.125 ≤ α ≤ 8. The correlations showed a mean deviation of less than 3% and a maximum deviation of less than 8.3% from the numerical data.