Convective heat transfer in open cell metal foams

Convective heat transfer in aluminum metal foam sandwich panels is investigated with potential applications to actively cooled thermal protection systems in hypersonic and reentry vehicles. The size eects of the metal foam core are experimentally investigated and the eects of foam thickness on convective transfer are established. Four metal foam specimens are utilized with a relative density of 0.08 and pore density of 20 ppi in a range of thickness from 6.4 mm to 25.4 mm in increments of approximately 6 mm. An exact-shapefunction nite element model is developed that envisions the foam as randomly oriented cylinders in cross o w with an axially varying coolant temperature eld. Our experimental results indicate that larger foam thicknesses produce increased heat transfer levels in metal foams. Initial FE simulations using a fully developed, turbulent velocity prole show the potential of this numerical tool to model convective heat transfer in metal foams. Metal foam sandwich panels have been proposed as alternative multi-functional materials for structural thermal protection systems in hypersonic and re-entry vehicles 1 . 2 This type of construction oers numerous advantages over other actively cooled concepts because of the unique properties of metal foams. These materials, when brazed between metallic face sheets, are readily suited to allow coolant passage without the addition of alien components that may compromise structural performance. Moreover, the mechanical properties can be varied to suit dieren t structural needs by varying the foam relative density. From a heat transfer point of view, these materials have been shown to be exceptional heat exchangers primarily due to the increased surface area available for heat transfer between the solid and uid phases. The thermo-mechanical response of metal foam sandwich panels has been recently studied and characterized. 2 In particular, it has been shown that using air as coolant at sucien tly high velocities, the strain due to buckling of these structures under thermo-mechanical loads can be virtually eliminated. The implementation of these materials in thermal protection systems, however, requires that a proper heat transfer model exists that allows the coupling between the thermo-mechanical and heat transfer problems to be properly analyzed. In other words, it is necessary to understand how dieren t foam properties such as relative density, pore density, and foam thickness will aect the heat loads that this type of structural component can remove. Heat transfer in metal foams has been a subject of active research in recent years. Lu et al. 3 developed an analytical model envisioning the foam as an array of mutually perpendicular cylinders subjected to cross-o w. In this study, a closed-form expression for the convective coecien t of a foam-lled channel with constant wall temperatures was presented based on foam geometry and material and uid properties. These authors reported that the simplifying assumptions used in their analysis were likely to lead to an over-prediction of the actual heat transfer level. This model has been partially validated by Bastawros and Evans 4 who performed forced convection experiments on aluminum foams adhered to silicon substrate face sheets. These authors reported that the predictions of Lu et al. 3 regarding the dependence of the convective coecien t on coolant velocity and strut diameter were qualitatively consistent with their observations, but that the foam thickness eects were not adequately modeled. In particular, they reported that the heat dissipation rate