PhD Defense | Edoardo Cascioli | Numerical analysis of low-Prandtl jets in turbulent forced convection regimes

Numerical analysis of low-Prandtl jets in turbulent forced convection regimes

The current need to ensure an effective and prompt transition of the energy sector towards zero-carbon has renewed the interest for nuclear technology. Small Modular Reactors (SMRs) seem particularly interesting for their reduced capital cost, operational flexibility and enhanced safety and security. Different SMR concepts are being developed around the world and the Liquid Metal Fast Reactor (LMFR) technology is one of the most convincing design options. In comparison with existing thermal reactors, the main advantages of LMFRs are the efficient use of nuclear fuel, possibility of partitioning and transmutation of long-lived fission products, inherent safety features and proliferation resistance, but also the enhanced heat transfer. The latter is directly linked to the high thermal conductivity of liquid metals, which, when expressed in dimensionless form, leads to these liquid metals having a very low (≪ 1) Prandtl number (Pr). These low Pr numbers necessitate some adjustments in Computational Fluid Dynamics (CFD) tools and methods, when predicting Turbulent Heat Transfer (THT) and thermal patterns. Within this context, the present doctoral research aimed to contribute to the testing and development of advanced low-Pr THT models for industrial CFD applications. In mixing pool-type LMFR facilities, such as the Multipurpose hYbrid Research Reactor for High-tech Applications (MYRRHA), under design at the Belgian Nuclear Center (SCK CEN), wall-unconfined forced convection flows, particularly free jets, of liquid metals play a fundamental role. Therefore, we selected low-Pr forced convection jet flows as the fundamental test case for our analysis. We performed Reynolds Averaged Navier-Stokes (RANS) and Transient-RANS (T-RANS) simulations for the direct testing of low-Pr THT models, but also Large-Eddy Simulations (LES) and combined LES-Direct Numerical  Simulations (LES/DNS) to generate numerical database of reference for validation purposes. LES/DNSs are also recommended to get the most accurate prediction of temperature fluctuations, which play a crucial role in the correct estimation of fatigue effects on structural materials. For full-scale reactor simulation and design studies, such high-fidelity CFD techniques are still too computationally demanding, and hybrid  (T-)RANS/LES techniques could be an interesting alternative to be studied as a follow-up of the present doctoral research. 

Promotor:

• Sasa Kenjeres (TU Delft)

Co-promotor:

• Chris Kleijn (TU Delft)

SCK CEN mentors:

• Katrien Van Tichelen
• Steven Keijers

Click here for a list of obtained PhD degrees.