Search

A CFD analysis has been carried out to study the thermal-hydraulic behavior of liquid metal coolant in a fuel assembly of triangular lattice. In order to obtain fast and accurate results, the isotropic two-equation RANS approach is often used in nuclear engineering applications. A different approach is provided by Non-Linear Eddy Viscosity Models (NLEVM), which try to take into account anisotropic effects by a nonlinear formulation of the Reynolds stress tensor. This approach is very promising, as it results in a very good numerical behavior and in a potentially better fluid flow description than classical isotropic models. An Anisotropic Shear Stress Transport (ASST) model, implemented into a commercial software, has been applied in previous studies, showing very trustful results for a large variety of flows and applications. In the paper, the ASST model has been used to perform an analysis of the fluid flow inside the fuel assembly of the ALFRED lead cooled fast reactor. Then, a comparison between the results of wall-resolved conjugated heat transfer computations and the results of a decoupled analysis using a suitable thermal wall-function previously implemented into the solver has been performed and presented.

The Loss of Coolant Accident in a Vacuum Chamber is one of the safety issues for the fusion facilities. With the aim of validating computer codes able to simulate thermal-hydraulic transient events in fusion reactors, experimental tests were carried out for comparing experimental data with simulations results. Among others, the Integrated ICE test facility (JAERI) has been developed to reproduce the ITER reactor geometry with a scaling factor of 1/1600. In order to perform safety assessment of fusion reactors, several computer codes have been developed and among them CONSEN, a fast running computer code developed to simulate thermal-hydraulic transients. The present paper is aimed at identifying the CONSEN capability to reproduce the behavior of main physical parameters during an Ingress of Coolant Event and consequent loss of vacuum in the Plasma Chamber of a fusion reactor. In particular, the CONSEN simulations results are here compared with experimental data obtained at the modified Integrated ICE test facility during the tests P1 and P7. Simulation results and experimental data have been compared and a good agreement has been identified for most of analyzed parameters. © 2015 Elsevier B.V. All rights reserved.

For the future development of Generation IV nuclear reactors, both safety and economic targets have to be achieved. In order to increase, at the same time, the power density generation and the safety features, a huge R&D effort is still required. Referring especially to Liquid Metal Cooled Fast Reactors, much attention is placed on Gas Entrainment (GE) phenomena, which could cause unlikely positive reactivity insertion accident. The GETS experimental facility (Gas Entrainment Test Section), especially aimed at studying the free surface vortices occurrence, has been built in the thermal-hydraulics laboratory of the DIAEE. The main purpose of this facility is to identify the most important parameters affecting the whirlpools formation and evolution. Experimental tests and preliminary observations have been performed. Different vortex behaviours related to different experimental conditions have been identified and presented in the present paper. 2D occurrence maps as function of different dimensionless groups (Reynolds, Froude and Weber numbers and H* = H/d ratio) have been defined. In the present paper, the results of a first experimental campaign, carried out with tap water, are discussed.

In the frame of the future generation of nuclear reactors, liquid metals are foreseen to be used as a primary coolant. Liquid metals are characterized by a very low Prandtl number due to their very high heat diffusivity. As such, they do not meet the so-called Reynolds analogy which assumes a complete similarity between the momentum and the thermal boundary layers via the use of the turbulent Prandtl number. Particularly, in the case of industrial fluid-dynamic calculations where a resolved computation near walls could be extremely time consuming and could need very large computational resources, the use of the classical wall function approach could lead to an inaccurate description of the temperature profile close to the wall. The first aim of the present study is to investigate the ability of a well- established commercial code (ANSYS FLUENT v.14) to deal with this issue, validating a suitable expression for the turbulent Prandtl number. Moreover, a thermal wall-function developed at Universite Catholique de Louvain has been implemented in FLUENT and validated, overcoming the limits of the solver to define it directly. Both the resolved and unresolved approaches have been carried out for a channel flow case and assessed against available direct numerical and large eddy simulations. A comparison between the numerically evaluated Nusselt number and the main correlations available in the literature has been also carried out. Finally, an application of the proposed methodology to a typical sub-channel case has been performed, comparing the results with literature correlations for tube banks.