An adapted steady RANS RSM wall-function for building external convection

•Steady RANS hcw predictions are studied on the front and rear faces of cubes.•RkE and RSM-LRNM models are accurate but standard TWF overestimate hcw.•An adapted TWF for the RSM is proposed based on the methodology of (Defraeye, 2011).•Customized TWFs perform satisfactorily for high z*.

Computational Fluid Dynamics (CFD) can improve usual estimates of building external convective heat transfer coefficients (hc,w) by accounting for the geometry of constructions, the aerodynamic field around them, the nature of convection and providing high resolution data. However, the limitations of usual mass flow descriptions and near wall treatments make the accurate prediction of hc,w challenging. Hence, this paper evaluates the ability of steady RANS Reynolds Stress (RSM) and k–ε realizable models to predict hc,w in case of isolated cubical obstacles. The accuracy of usual CFD methods and turbulence models, as well as fine grid near wall models and usual temperature wall functions (TWFs) are examined by comparison with experimental and detailed numerical data.When used with a low Reynolds number model (LRNM), both turbulence models accurately predict hc,w on the front and rear faces of the obstacle. However, they show different behaviors on the other faces and highlight issues related to the dynamic behavior of real flows. Moreover, hc,w predictions obtained using standard TWFs substantially deviate from the validated LRNM results. Therefore, a customized TWF suited for use with the RSM and forced convection problems is proposed by extending studies of Defraeye et al. (An adjusted temperature wall function for turbulent forced convective heat transfer for bluff bodies in the atmospheric boundary layer; Building and Environment, 2011, 46, 2130–2141). Customized TWFs substantially improve WF-based hc,w predictions with respect to LRNM results while keeping their cost effectiveness, and provides satisfactory results even for high z∗.