Numerical Determination of Convection Coefficients for Internal Surfaces in Buildings Dominated by Thermally Stratified Flows

Convection coefficients for internal building surfaces, characterized by the air-surface temperature difference and local airflow, are crucial to a reliable building energy simulation and analysis based on the integrated modeling of heat and air flow in buildings, especially dominated by the thermally stratified flows, where the coefficients may vary significantly along each vertical wall. A turbulence model with a differential viscosity and dynamic turbulent Prandtl is proposed to predict the convective heat transfer for the case of thermally stratified flows in buildings. The local convection coefficients and then the bulk ones are numerically derived based on the law-of-the-wall method. Computational fluid dynamics (CFD) simulations are performed in an experimental enclosure and the results of air temperature profiles in the space are compared with the experimental data. Good agreements of thermally stratified pattern between the CFD and experiments are obtained. Although some discrepancies in the bulk convection coefficients are observed when natural convection is dominating, little differences in the temperature profiles give much confidence. As a whole, the CFD method using the turbulence model suggested in this study exhibits its good ability to effectively predict the convection coefficients.