MELTING OF A SOLID ADJACENT TO A HEATED VERTICAL CYLINDER WITH OR WITHOUT SUBCOOLING OF THE SOLID

Experiments were performed to provide quanitative heat transfer data corresponding to the problem of melting about a heated vertical cylinder embedded in a solid phase-change material. The phase change material employed was 99 percent pure n-eicosane paraffin having an experimentally determined melting point of 36.4 C. Experiments were conducted with the solid phase either at the melting point or subcooled by 14.4 C.;The ranges of power density at the heated cylinder surface varied from 175 to 3080 W/m('2) and 506 to 3034 W/m('2) for the non-subcooled and subcooled tests respectively. These power ranges produced respective Rayleigh number ranges in the melt of 1.07 x 10('9) to 1.10 x 10('10) and 2.55 x 10('9) to 1.22 x 10('10), with the characteristic length in the Rayleigh number being the height of the heated tube.;Measured transient cylinder-wall temperatures demonstrate that the present melting problem is characterized by a conduction heat transfer regime at early times followed by a transition to a natural convection dominated regime. During the later stages of the natural convection regime, the cylinder wall temperatures reach a steady state condition.;Transient heat transfer coefficients were calculated for each test. Comparison of the coefficients for tests employing a no-slip upper-surface velocity boundary condition to those for a slip condition demonstrate that the change in the upper-surface velocity boundary condition has no effect on the heat transfer coefficient. Comparisons between tests initiated with and without subcooling show that the impact of an initially subcooled solid is to greatly delay the onset of the natural convection regime.;The paraffin is contained within a large cylindrical vessel which, in turn, resides in a constant temperature water bath for purposes of thermal control. The cover assembly of the containment vessel provides for either a no-slip or slip velocity boundary condition along the upper bounding surface of the paraffin. Situated vertically along the central axis of the vessel is an electrically heated cylinder 2.54 cm in diameter, and having a height to diameter ratio of 10. The wall of the cylinder is instrumented with thermocouples which provide local temperature data.;The steady-state regime was defined to exist once the measured heat transfer coefficient had attained values within five percent of the final value. In general, the net energy input to the paraffin was found to be greater for the subcooled versus the non-subcooled tests prior to the onset of the steady-state regime. Furthermore, it was found by a comparison of steady-state values for tests initiated with and without subcooling that an initially subcooled solid reduces the magnitude of the natural convection heat transfer coefficient by about ten percent.;The steady-state heat transfer coefficients for non-subcooling and subcooling tests were separately correlated. An initial correlation for each case employed the Nusselt and Rayleigh numbers, with the characteristic length in each parameter being the heated-cylinder height. A second correlation for each of the two cases included, in addition to the Nusselt and Rayleigh numbers, a heated-cylinder height to melt layer thickness ratio. The melt layer thickness was also employed as the characteristic length in the Nusselt and Rayleigh numbers. The correlations for the non-subcooling tests are in good agreement with others corresponding to natural convection in vertical spaces. The correlations for the subcooling tests are not in good agreement indicating the presence of other participating transfer processes.;Melt shapes corresponding to the completion of selected tests were measured and are reported. The inverted bell shape of the cavities demonstrates the dominant role of natural convection in the melting problem. Melt cavities for the subcooling tests were, as expected, smaller than their counterparts initiated without subcooling.