Heat transfer coefficient during flow boiling in a minichannel at variable spatial orientation

Abstract

The paper presents flow boiling heat transfer of FC-72 in 1mm×40mm×360mm minichannel (depth×width×length) where foil with evenly distributed micro-recesses acts as a heating surface. The minichannel is set at various angles, i.e. 0°, 45°, 90°, 135° and 180° inclinations in relation to the horizontal plane. Due to liquid crystal thermography it is possible to determine the two dimensional temperature of the plain side of the foil. The one- and two-dimensional heat transfer approaches are proposed to calculate the heat transfer coefficient. In the two-dimensional approach the inverse problem in the heating wall is solved by the semi-analytical method based on Trefftz functions in subareas, abbreviated to the nodeless Trefftz method. Local values of heat transfer coefficient on the surface between the heating foil and boiling liquid were calculated on the basis of the third-kind boundary condition. It turned out that the orientation of the minichannel has a significant impact on heat transfer coefficient. The values of heat transfer coefficient were analyzed separately for the subcooled and saturated nucleate boiling regions. The highest values of the heat transfer coefficient were observed in the saturated boiling region – for the vertical minichannel, position 90° and in the subcooled boiling region – for the inclined minichannel, position 45° (although for the vertical minichannel, position 90° heat transfer coefficients were also high). The lowest values of the heat transfer coefficient for both regions of boiling were obtained for the horizontal minichannel, position 180° but for the subcooled boiling region very low heat transfer coefficients were obtained also for the inclined minichannel, position 135°. The same tendencies in the results were observed for the two applied methods. The polynomial smoothing of the results in the 2D method can affect the rounding values of heat transfer coefficient. The one-dimensional approach seems to be less sensitive to measurement errors.