| Compact heat exchangers have been designed following the same basic methodology for over fifty years. However, with the present emphasis on energy efficiency and light weight of prime movers there is increasing demand for completely new heat exchangers. Moreover, new materials and mesoscale fabrication technologies offer the possibility of significantly improving heat exchanger performance over conventional designs. This work involves fundamental flow and heat transfer experimentation to explore two new heat exchange systems: in Part I, large arrays of impinging jets with local extraction and in Part II, metal foams used as fins.Jet impingement cooling is widely used in applications ranging from paper manufacturing to the cooling of gas turbine blades because of the very high local heat transfer coefficients that are possible. While the use of single jet impingement results in non-uniform cooling, increased and more uniform mean heat transfer coefficients may be attained by dividing the total cooling flow among an array of smaller jets. Unfortunately, when the spent fluid from the array's central jets interact with the outer jets, the overall mean heat transfer coefficient is reduced. This problem can be alleviated by locally extracting the spent fluid before it is able to interact with the surrounding jets. An experimental investigation was carried out on a compact impingement array (Xn/Djet = 2.34) utilizing local extraction of the spent fluid (Aspent/Ajet = 2.23) from the jet exit plane. Spatially resolved measurements of the mean velocity field within the array were carried out at jet Reynolds numbers of 2300 and 5300 by magnetic resonance velocimetry, MRV. The geometry provided for a smooth transition from the jet to the target surface and out through the extraction holes without obvious flow recirculation. Mean Nusselt number measurements were also carried out for a Reynolds number range of 2000 to 10,000. The Nusselt number was found to increase with the Reynolds number to the 0.6 power with peak Nusselt numbers near 75 at a Reynolds number of 10,000.Open-celled metallic foams offer three important characteristics which enable them to perform well in heat exchange applications. They contain a very large surface area to volume ratio, a highly complex flow passage through the foam, and in many cases, significant thermal conductivity in the solid phase. Unfortunately, difficulty arises when metal foams are implemented in heat exchanger designs. The performance of the foam has not been characterized in a way which is conducive to analytical design of high performance heat exchangers. The second part of this work provides both flow and heat transfer measurements for metal foam geometries. Full-field velocity measurements through a foam sample were acquired using MRV. The measurements show transverse velocities on the order of 25-30% of the Darcy velocity, UD, which produce enhanced thermal dispersion within the foam matrix. A mechanical dispersion coefficient, DM, was formed which demonstrates the transverse dispersion to be 13 times the kinematic viscosity and 9 times the thermal diffusivity of air at 20°C and 1 atm. To describe the heat transfer performance of the foam as a fin, we have developed a new method that utilizes a well documented, periodic heat exchanger core test and a new one heated wall (OHW) test which when used in conjunction are shown to determine the convective performance (hmAc), the conductive performance (ksAc), and the effective bond resistance associated to attaching metal foams to primary heat transfer surfaces (RBond). Small pore diameter foams, d &le 1 mm, where found to perform approximately a factor of 2 greater per unit volume than a comparable fine-fin heat exchanger surface at the same pumping power which points to the fact the foam as a system is conduction limited not convection limited. |
