Abstract
Selected results of three Direct Numerical Simulations are presented, on relevant test cases for the thermal hydraulics of liquid–metal-cooled nuclear reactors, encompassing a wide spectrum of turbulent convection regimes. The first test case is a Rayleigh-Bénard cell at a Grashof number Gr=5×107, representative of the conditions in the unstably stratified layer of coolant in a reactor pool in both standard operating conditions and emergency situations, e.g. shutdown of the cooling system. The second case is the mixed convection in a cold-hot–cold triple jet configuration, representative of liquid streams exiting from the core into the pool, and relevant for the modeling of thermal striping and thermal fatigue phenomena on the vessel containment walls. The third case is the fully-developed flow in a vertical bare rod bundle with triangular arrangement and a pitch-to-diameter ratio P/D=1.4, in both forced and mixed convection conditions. These regimes respectively represent normal operation or decay heat removal conditions in reactor cores. The availability of these numerical databases will allows for an in-depth analysis of the turbulent flow and heat transfer in liquid metals under different convection regimes, and is also relevant for the development, calibration and validation of turbulent heat transfer models.
Highlights
Convection heat transfer in liquid metals is a fundamental topic of relevant interest in the field of nuclear engineering
Since the execution of detailed experiments is hampered by practical difficulties which can be overcome only with the design of specific experimental setups and procedures [3], the resort to highly resolved Direct Numerical Simulation is a viable alternative to provide reliable data for the development, calibration and validation of advanced turbulent heat transfer models, given the accuracy of models for nuclear applications based on the Simple Gradient Diffusion Hypothesis (SGDH) and the specific behavior at low Prandtl number is still a matter of active debate [4,5,6]
Three Direct Numerical Simulations (DNSs) are presented with the aim of providing a database of statistics which may be useful for the development and validation of turbulent convection models specific for liquidmetal-cooled reactors
Summary
Convection heat transfer in liquid metals is a fundamental topic of relevant interest in the field of nuclear engineering. Since the execution of detailed experiments is hampered by practical difficulties which can be overcome only with the design of specific experimental setups and procedures [3], the resort to highly resolved Direct Numerical Simulation is a viable alternative to provide reliable data for the development, calibration and validation of advanced turbulent heat transfer models, given the accuracy of models for nuclear applications based on the Simple Gradient Diffusion Hypothesis (SGDH) and the specific behavior at low Prandtl number is still a matter of active debate [4,5,6]. In case of use of a SGDH model, the selection of the most suitable turbulent Prandtl number is the main difficulty in lowPrandtl number representations of turbulent heat transfer phenomena. Instead, when a more detailed study is conducted, it should be considered that large-scale fluctuations are in general smaller for order one molecular Prandtl number with respect to the low-Prandtl number fluids, see [9]
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
