Solar Energy Laboratory

Research > Projects

Natural convection in metal foams


The objective of this project is to evaluate the use of metal foams and foam-like structures to enhance energy transfer rates to immersed heat exchangers in water storage. The goal of thermal design is to achieve the largest energy transfer rate with the least material, and at reasonable cost. Metal foams are considered potentially beneficial because they have a large heat transfer area/volume ratio and low specific weight. Metal wire mesh as a lower cost option.

The governing heat transfer laws in natural convection for foam/water systems have been determined from experiments with copper and reticulated vitreous carbon (RVC) foams and stacks of copper mesh screens. (The RVC foam was evaluated to extend the data base. It does not serve as an effective heat transfer medium because of it low thermal conductivity.) For foam, two types of heated systems were considered: a layer of foam fully saturated with water and a layer of foam with a free layer of water above or inter-posed between two foam layers. In both cases, the system is heated at the lower surface and cooled at the upper surface, thus setting up natural convection.

Fig. 1. Copper foam

A single heat transfer correlation of Nusselt number was obtained in terms of Rayleigh number and a modified Prandtl number which relates fluid properties and hydrodynamic effects in the foam when fluid motion is present. The developed heat transfer correlation provides the ability to relate the structural properties of the foam (density and pore size) to the degree of enhancement of energy transfer rate. Conduction is the primary energy transfer mechanism. As a consequence, the copper foam enhanced heat transfer compared to water alone. The “foam-like” structure made of commercial copper screens does not provide the same benefit as the copper foam. There are two reasons for the poor performance. The stagnant thermal conductivity is about one-fourth that of the copper foam because of poor thermal contact between the screen layers in the stack. In addition, the permeability of the screen layers is lower than the foam and thus the convective motion is constrained. Future efforts in our laboratory will consider more porous foam like structures with less contact resistance.

Our results suggest that to be effective in this application, foams must be highly porous and made of conductive metals. For practical applications, there exists an optimal thickness of foam adjacent to the primary heat transfer surface, and there would be a diminishing advantage and cost penalty for adding foam beyond it. On-going work seeks to identify the optimal thickness and develop heat exchanger designs.