Page 44 - Thermal Hydraulics Aspects of Liquid Metal Cooled Nuclear Reactors

Thermal-hydraulic challenges in liquid-metal-cooled reactors 19

2.3 Categorization

‘In order to structure the information gathered from the identification process, the chal-
lenges have been divided in three main categories: core thermal hydraulics, pool thermal
hydraulics, and system thermal hydraulics. For each of these main categories, a division
is made between normal reactor operation, off-normal conditions, and severe accidents.
Seven basic phenomena that are at the basis of the challenges above and that
involve prenormative and fundamental challenges have been classified in a separate
category. These seven basic phenomena are turbulent heat transfer, thermal fluctua-
tions, mechanical fluctuations, mass transfer, bubble transport, particle transport, and
solidification.
Finally, a fifth category was added that deals with guidelines for numerical simu-
lation and experiments.

2.4 Thermal hydraulic challenges

2.4.1 Basic phenomena

The seven basic phenomena that were identified are the following:
l Turbulent heat transfer (see also Section 6.2.1)
Challenge
Simulation of turbulent heat transfer in highly conductive liquid metals (low-Prandtl-number
fluids) is a challenge for CFD because of the large difference in boundary layers for the
momentum and energy transport (Roelofs et al., 2015a).
State of the art
Today, advanced models for industrial CFD simulation exist but are in general only valid
in one flow regime (i.e., natural or forced convection). Roelofs et al. (2015a) summarize
the development status of different modeling approaches. The most promising one is
described in more detail by Shams et al. (2014) who describe a model that for some basic
test cases operates well under both forced and natural convection conditions. Shams
(2017) presents an update of this model for high Rayleigh number natural convection
regimes.
Development needs
As explained by Roelofs et al. (2015b), the ultimate goal is to have a RANS turbulent heat
transfer model that is valid in all flow regimes (natural-mixed-forced) simultaneously avail-
able in at least one of the major engineering CFD codes (i.e., commercial codes or open-
source codes like OpenFOAM and Code Saturne). The model presented by Shams et al.
(2014) and updated by Shams (2017) is the most advanced model in this respect, but to
develop it further, new reference data, both experimental and high-fidelity numerical, have
to be generated. For this purpose especially, new data in the mixed convection regime and
data for flow separation and nonconfined flows are being generated and should be used in the
near future to further develop and validate the necessary models. Similar goals are being
pursued with an emphasis on hybrid RANS/LES turbulence modeling within the spectral
element code Nek5000 (Bushan et al., 2018).