COMPUTATIONAL METHODS AND TURBULENCE IN MODELING IN TURBOMACHINERY

摘要

The computational fluid dynamics is a powerful tool for the study of flow in industrial cases. With the exception of the simulation methods DNS and LES, the modeling techniques based on the Reynolds Averaged Navier Stokes equations are still the most cost effective solution for an industrial problem related to fluid mechanics. CFD has used so far to obtain accurate predictions for complex internal and external flows. The accuracy has been well supported by the advanced turbulence models and the available CPU power that gives the ability to proceed to predictions with well-formed grids in large problems. The modeler has the chance to select among a plethora of discretization techniques, turbulence models and programming architectures in order to achieve accurate and realistic predictions. Unfortunately, there is no one standard choice capable to fit to all the modeling problems. It is the modeler's experience that plays an important role in the choice of a certain technique appropriate to a certain problem, since most of these techniques have been developed and calibrated "to work" always on specific problems. In this work, a presentation of the most widely used CFD techniques, turbulence models and programming environments will be attempted.