Elements of a General Correlation for Turbulent Heat Transfer

Abstract

Typically, heat transfer researchers present results in the form of an empirically based relationship between a length-based Nusselt number, a length-based Reynolds number, and a fluid Prandtl number. This approach has resulted in a multitude of heat transfer correlations, each tied to a specific geometry type. Two recent studies have contributed key ideas that support the development of a more general correlation for turbulent heat transfer that is based on local parameters. Maciejewski and Moffat (1992a, b) found that wall heat transfer rates scale with streamwise turbulent velocity fluctuations and Anderson and Moffat (1992a, b) found that the adiabatic temperature rise is the driving potential for heat transfer. Using these two concepts and a novel approach to dimensional analysis, the present authors have formulated a general correlation for turbulent heat transfer. This correlation predicts wall heat flux as a function of the turbulent velocity fluctuations, the adiabatic temperature rise, and the fluid properties (density, specific heat, thermal conductivity, and viscosity). The correlation applies to both internal and external flows and is tested in air, water, and FC77. The correlation predicts local values of surface heat flux to within ± 12.0 percent at 95 percent confidence.