Page 52 - Thermal Hydraulics Aspects of Liquid Metal Cooled Nuclear Reactors

Thermal-hydraulic challenges in liquid-metal-cooled reactors 27

free surface, and gas entrainment (Tenchine, 2010; Chellapandi and Velusamy, 2015).
Whereas for the individual challenges a high-resolution experimental and/or numerical
approach is often feasible, the large ranges of spatial and time scales present in the complex,
three-dimensional reactor plena require a balanced choice between resolution and cost of
experiments and simulations.
State of the art
A number of scaled integral experiments, typically employing water as simulant fluid,
have been performed in the past to study the flow behavior in the upper and lower plena of
liquid-metal reactor pools (Roelofs et al., 2013; Planquart and Van Tichelen, 2017). New
experiments are in operation for the reactor designs that are being developed today. Water
is still mainly used as simulant fluid as it allows the use of high-resolution optical techniques
for the characterization of velocity (and even temperature) fields (Gu enadou et al., 2015;
Planquart and Van Tichelen, 2017). Scaled experiments with liquid-metal coolant are scarce
(Tarantino et al., 2015; Tarantino, 2017; Van Tichelen and Mirelli, 2017). Because of the lim-
itations of the essentially one-dimensional system codes in modeling the phenomena in the
three-dimensional reactor plena, CFD codes have been developed for this purpose since the
1980s. Nowadays, these codes allow the 3-D modeling of the reactor plena in steady-state
and transient conditions, with certain simplifying approaches such as the use of a porous mesh
for complicated structures and with reasonable results with respect to temperature gradients
and fluctuations as a result of limitations in turbulence modeling.
Development needs
Development and validation of integrated models on the basis of properly scaled integral
experiments is absolutely necessary. With the increasing capabilities of CFD, an increased
level of instrumentation in experiments is requested, providing the detailed data necessary
for validation of a CFD approach. Up to now, limited but valuable data are available from
facilities employing liquid metals. Additional efforts are to be made here, including the
development of instrumentation that provides a good characterization of the flow field in
opaque liquids. Apart from that, further development and validation of fast reduced-order
numerical approaches such as grid-free methods (e.g., Prill and Class, 2014) should be envis-
aged. Such reduced-order modeling for thermal stratification and other pool dynamics
effects is being pursued momentarily. Coarse-grid CFD and proper orthogonal decomposi-
tion (POD)-based modeling are actively being developed within the system code SAM
(Hu, 2017) and Nek5000 (Merzari et al., 2017).
Thermal stratification and fatigue (see also Section 6.2.4)
l
Challenge
In liquid-metal pools, the inertial and buoyancy forces are of similar order of magnitude. As
a result, thermal stratification can occur in which a high-temperature layer separates from a
low-temperature layer in the vertical direction. The interface is characterized by a large tem-
perature gradient and unstabilities that are passed on to the surrounding structures, possible
leading to low-cycle thermal fatigue. Hence, mitigation of stratification is an important
objective in reactor design.

State of the art
Thermal stratification has been studied extensively for reactor design purposes in dedicated
experimental setups using water and, to a lesser extent, sodium (Kimura et al., 2010;
Tenchine et al., 2012). Conditions for stratification to occur were established as a function
of Richardson and Peclet numbers. The old and new experimental programs on pool thermal
hydraulics (see above) often inherently also address thermal stratification and thermal
fatigue issues. Numerical modeling of stratification asks for CFD techniques with locally