• Pressurized Thermal Shock
• Turbulence
• CFD
• ROCOM

Pressurized Thermal Shock (PTS) has been identified as one of the most important industrial issues related to nuclear reactor safety, because of the high criticality of this accident scenario. Therefore, assessment of PTS is important for long term operations. A reliable determination of the thermal loads requires accurate prediction of the mixing phenomena in the cold leg and downcomer. In reality, strong mixing takes place, and the detailed mixing pattern is required to make accurate and realistic predictions. This pattern is the consequence of a complex three-dimensional fluid flow, where density differences between the coolant water and the primary loop inventory can play an important role, inducing buoyancy-driven mixing. For these reasons CFD codes can play an important role in the analysis of the PTS scenario, altough it is generally stated that the existing models for the turbulent mixing phenomena should be improved. It has been supposed that these errors could derive from the limitations of RANS in modelling the complex buoyancy flow and, consequently,it was suggested the study with advanced CFD methods, like LargeEddy Simulation. In fact, LES application to single-phase PTS showed a good qualitative agreement. The main issue with this kind of models is, however, the really high computational cost that make them of hard use for industrial aims. Since several progresses have been made in the improvement of RANS models it became important to understand if these models could replace LES with an acceptable loss of accuracy that would be repaid with an high gain in computational cost. The objective of this work is to compare results of simulations for different coolant mixing scenarios performed with new RANS models with data obtained from experiments conducted in ROCOM and UPTF test facilities and results from LES simulation.