On the adequacy of wall functions to predict condensation rates from steam-noncondensable gas mixtures

As one looks forward to applying CFD based methods to simulate turbulent flows in larger volumes up to containment scales, the mesh resolution, especially near the walls, becomes one of the main issues dictating the feasibility of the simulation. The wall-function approach is a natural choice to minimize the computational size of the problem and make it tractable. In the current investigation, we compare the wall-function to the fully resolved boundary layer approaches for the prediction of vapor condensation rates on cold walls in the presence of noncondensable gases. We simulate three sets of geometric configurations. The first two sets relate to domains which are small (height of 2 m) and medium (height 4.8 m), and for which experimental heat transfer data are available. In the third set, we look at a hypothetical large 2D rectangular domain in which the condenser height is comparable to that of typical NPP containments (20 m). developing region of the boundary layer, it is found that the wall function treatment leads to substantial deviations from the wall resolved approach and available experimental data. Further downstream, however, when the boundary layer is fully developed, the discrepancy is greatly reduced. therefore concluded that the wall-function formulation is able to provide predictions of condensation rates that are similar to wall-resolved treatments in simple forced flows for which fully developed boundary layers can be assumed over most of the domain. Care must however be exercised to ensure the chosen wall functions relate to the flow regime under consideration.