| In order to extend LES to flows of engineering interest, wall-models are required, because the resolution of the near-wall layer is highly demanding in terms of grid-points. In this study the Two-layer model is studied. The Two-layer model is based on a zonal approach in which the filtered Navier-Stokes equations are solved far from the wall, while in the near wall-layer the boundary layer equations are solved and turbulence is modeled using an algebraic turbulence model. The underlying hypotheses on which this model is based are much more general and less restrictive than other models presented in literature.; Tests for simple flows, such as channel and zero-pressure gradient boundary layer flows, showed good results at a fraction of the cost of a resolved calculation, however the few tests in the literature, over more complex flows such as a backward-facing step or boundary layers with pressure gradients, were unsatisfactory.; Algebraic turbulence models perform poorly in complex flow, thus, to extend the Two-Layer Model to more complex flows a more advanced turbulence model has been implemented.; The new Two-Layer Model has been validated in channel and zero-pressure gradient boundary-layer flows. Tests over the backward-facing step at low and high Reynolds numbers, showed excellent results compared to the reference data, much better than those reported in the literature for the same test case. In particular a better accuracy and robustness have been achieved in comparison to the algebraic turbulence model. Tests for a favorable pressure gradient boundary layer, however, were less accurate.; The cost of the Two-Layer Model is less than 5% of the total cost of the calculation, and the grid for a simulation requires more than one order of magnitude less points than a resolved calculation. The cost of the new Two-Layer Model is about 10 ÷ 15% more than the original algebraic model. |
