Improving predictions of heat transfer in indoor environments with eddy viscosity turbulence models

Heat transfer modelling in indoor environments requires an accurate prediction of the convective heat transfer phenomenon. Because of the lower computational cost and numerical stability, eddy viscosity turbulence models are often used. These models allow modification to turbulent Prandtl number, and near wall correction which influences stagnation points, entrainment, and velocity and time scales. A modified v²-f model was made to correct the entrainment behaviour in the near wall and at the stagnation point. This new model was evaluated on six cases involving free and forced convection and room airflow scenarios and compared with the standard k-ε, and k-ω-SST models. The results showed that the modification to the v²-f model provided better predictions of the buoyant heat transfer flows while the standard k-ε failed to reproduce and underestimate the convective heat transfer. The k-ω-SST model was able to predict the flow field well only for a 2D square cavity room, and 3D partitioned room case, while it was poor for the other four cases.