Development and evaluation of advanced turbulence models for high Reynolds number separated flows

One of the limiting factors in current industrial Computational Fluid Dynamics is the way large-scale structures, that might not be turbulent, are represented by methods based on statistical models (RANS). The most accurate method available today for industrial applications (Large-Eddy simulation, LES) still remains largely unpractical (in terms of cost, but also in terms of physics that need to be modeled). A method combining both approaches (Hybrid RANS/LES) has been developed in Simcenter STAR-CCM+ (a leading CFD software, developed by Siemens PLM) and the aim of the project is to improve this methodology (originally derived and validated on simple, canonical, low Reynolds number cases) for high Reynolds number separated flows. In particular, the focus will be on the optimization of offshore platforms in order to increase their operational efficiency and to reduce the probability of incidents.