Conference Agenda

Progressing in a line of research related to the description of heat transfer phenomena for fluids at supercritical pressure as resulting from predictions obtained by CFD models, the present paper displays information about the distribution of fluid velocity, density, temperature and other relevant variables along and across vertical heated channels in upward flow. The main purpose of the work is to try describing in a meaningful way some of the phenomena observed in previous work by the Authors, in similarity with observations proposed in literature since decades, in order to contribute drawing an overall picture of flow and heat transfer regimes as it is possible to infer with some degree of fidelity by available CFD models.

Examples of the variety of fluid-dynamic and heat transfer phenomena presented by fluids at supercritical pressure that can be considered in similarity with the behaviour observed at subcritical pressure are first reported, drawing from previous work by the Authors. A first example is related to flow stability, for which similar phenomena can be observed with a striking similarity between the Ishii-Zuber plane representation, as a function of the phase-change and the subcooling numbers, and the corresponding one at supercritical pressure as a function of a trans-pseudocritical and a sub-pseudocritical numbers. A second example is related to the characterisation of a specific mode of deteriorated heat transfer, pointed out by the experimentally observed jumps in the wall temperature trends to a higher manifold when the phenomenon of flow laminarization occurs close to the heated wall, which allowed to understand the mechanism underlying the observed deteriorated heat transfer mode at sufficiently large heat flux making use of a pseudo-Nukiyama curve.

Despite of these similarities, subcritical and supercritical flow and heat transfer phenomena are anyway very different and deserve specific descriptions. In the aim to summarise in a meaningful way the lesson learned so far in relation to the observed heat transfer and flow regime phenomena in vertical upward flow at supercritical pressure, supercritical water operating conditions and supercritical carbon dioxide experiments addressed by CFD analyses in several recent works have been considered. Disregarding the original reasons for coping with these operating conditions, respectively related to the analysis of the effect of heating structures on flow stability predictions at supercritical pressure and to the comparison of the results of an algebraic heat flux turbulence model with experimental data, attention is here put to the coherent picture of phenomena provided in both cases by the analysis of flow regimes and heat transfer deterioration by the adopted computational fluid-dynamic (CFD) models.

It is remarked that the two fluids addressed in this work, water and CO₂, are respectively the reference fluid to be considered for applications related to SuperCritical Water-cooled Reactors (SCWR) and one of the surrogate fluids most used in performing experiments in correspondingly similar conditions. While in the former case we describe postulated phenomena occurring in water, having only a qualitative counterpart in known experimental observations, in the latter case the presented distributions are granted the support of a good comparison with experimental wall temperature data. The picture obtained for both fluids is sufficiently coherent to suggest a possibly general description of the most relevant observed phenomena.

With no pretence to be exhaustive, the work is intended propose a meaningful presentation of phenomena occurring at supercritical pressure in vertical upward heated channels, as a complement to and an extension at supercritical pressures of the well assessed body of knowledge available for flow and heat transfer regimes observed in boiling channels.

Wire-wrapped fuel assemblies are considered for heavy liquid-metal-cooled reactor concepts due to their compact spacer structure, enhanced mixing capability, and favorable thermal-hydraulic performance. However, the helical wire spacers generate complex three dimensional flow structures, which make the prediction of local velocity, turbulence, and temperature fields challenging. Moreover, thermal-hydraulic behavior is sensitive to geometric variations. In particular, fuel assembly deformation may modify the coolant redistribution and affect the prediction of local hot spots. Therefore, high-quality experimental data are required to support CFD (Computational Fluid Dynamics) validation and to reduce the uncertainty in the prediction of maximum cladding temperature.

Within the framework of the ANSELMUS (Advanced Nuclear Safety Evaluation of Liquid Metal Using Systems) project, this work contributes to the fuel assembly safety activities by generating experimental validation data for nondeformed and deformed wire wrapped fuel bundle configurations. The investigated geometry is a scaled 7-rod wire wrapped bundle representative of the MYRRHA fuel assembly. The main dimensionless geometrical parameters are preserved, including the pitch-to-diameter ratio and the wire axial pitch-to-diameter ratio. The bundle has a rod diameter of 10 mm, a wire diameter of 2.8 mm, a pitch-to-diameter ratio of 1.28, and a wire axial pitch-to-diameter ratio of 40.

The experiments are conducted in the AUPINEL (ANSELMUS & PASCAL INtegrated watEr Loop) closed-loop water facility at the von Karman Institute, using demineralized water as the working fluid. The metallic test section is equipped with transparent windows, providing optical access for flow and thermal measurements as well as internal access for deformed scenario. This design enables measurements under both nominal and locally deformed configuration within the same experimental setup.

The experimental campaign combines flow field and thermal field measurements. Particle image velocimetry (PIV) is used to obtain spatially resolved velocity fields at different axial locations and Reynolds numbers. There exist several measurement plane that are located in the gap between the hexagonal channel wall and the front rods. Velocity magnitude, turbulent kinetic energy, and vorticity fields are analyzed to assess the influence of wire position, axial location, deformation, and heating condition. The results show that the axial location has a strong influence on the flow field and that the helical wire locally redistributes the flow. The deformation element strongly modifies the velocity field at the insertion plane, while its effect on farther planes is weaker but still visible, mainly in the turbulence field. For the tested conditions, heating has a negligible influence on the velocity field, consistent with the low Richardson number.

Thermal measurements are performed using 15 surface thermocouples distributed over three axial locations, together with liquid crystal thermography on the visible rod surfaces. The results show that the local surface temperature depends strongly on the position relative to the wire and on the axial location. The deformation effect on temperature is generally limited, although more visible differences are observed at the top measurement plane.

Overall, this work provides a dedicated experimental database for nominal and deformed MYRRHA-representative wire wrapped bundle configurations. The combined pointwise and spatially resolved measurements support code-to-experiment comparison, CFD validation, and the development of improved modelling guidelines for thermal-hydraulic analyses of wire-wrapped fuel assemblies. Comparison with numerical results and lessons learned from this validation test case will be presented.

Two-phase flow phenomena are central to the thermal-hydraulic safety analysis of water-cooled nuclear reactors. For example, boiling heat transfer exhibits a variety of flow regimes governed by bubble coalescence and breakup dynamics. One such regime is slug flow, characterized by the formation of elongated bullet-shaped gas bubbles known as Taylor bubbles.

This study investigates the flow and material parameter range governing the evolution of capillary waves and interfacial instabilities in Taylor bubbles under counter-current laminar flow. The key dimensionless numbers, namely the Reynolds (Re), Capillary (Ca), and Eötvös (Eö) numbers, are systematically varied to quantify their influence on bubble shape and size, instantaneous bubble rise velocity, and the surrounding flow structure. Since capillary waves occur on length scales of several tens of micrometers, high-fidelity three-dimensional simulations are computationally prohibitive. Therefore, the present study employs two-dimensional axisymmetric simulations using both OpenFOAM and ANSYS Fluent. The computational domain consists of a vertical pipe of diameter D of 12.4 mm and length 20D. Sharp liquid-gas interface capturing is achieved by using two different methods: (a) geometric Volume of Fluid (VOF) method with Piecewise Linear Interface Construction (PLIC) using OpenFOAM and (b) a Coupled Level Set Volume of Fluid (CLSVOF) method using ANSYS Fluent. The two platforms are used for code-to-code comparison, and results are validated against experimental measurements from the THELMA lab.

Findings reveal that the tail instability is strongly governed by the inertial, viscous, and surface tension forces, with critical transitions occurring within specific ranges of Re, Ca, and Eö. In contrast, the nose of the bubble remains smooth and stable across the investigated parameter range. Moreover, the predicted liquid film thickness shows close agreement with the relevant experimental data. Simulations also indicate the presence of interfacial undulations near the bubble tail, resembling capillary wave formation; however, a more detailed study is required to fully characterize these undulations.

The interaction between a hot main flow and a colder, stagnant fluid within T-junctions gives rise to complex thermo-hydraulic phenomena, such as thermal stratification and temperature fluctuations. A thorough understanding of these mechanisms is essential for the design and operational assessment of piping systems across numerous industrial applications. While numerical simulations provide valuable insights, they require robust experimental data for validation, particularly concerning the nuanced behaviors that dictate thermal loading on the pipe structure.

This paper presents an experimental investigation aimed at characterizing and comparing thermal behavior within both vertical and horizontal dead-legs. The study employs a non-intrusive methodology based on high-resolution thermal imaging of the external pipe surface (DN100) to infer the internal fluid dynamics and thermal patterns. This approach allows for a detailed visualization of phenomenology without disturbing the flow. The presentation will focus on the influence of dead leg orientation on the thermal phenomena. The tests are performed for a flow velocity between 90 m3 /h and 140 m3 /h and a water temperature of 80°C.

Our observations reveal that phenomenology is highly dependent on dead-leg orientation and is consistent with findings in the literature. For the vertical dead-leg, the behavior is characterized by a "piston-like" effect, driven by the movement of a distinct stratification plug. The thermal field is divided into two primary zones: an upper zone near the main line, subject to large-amplitude temperature fluctuations, and a lower zone where these fluctuations are significantly attenuated. A detailed analysis shows that these two zones are correlated but exhibit a clear phase shift, indicating a complex energy transfer mechanism.

In contrast, the horizontal dead-leg exhibits a fundamentally different phenomenology. Here, the dominant mechanism is the formation of a "tongue-like" intrusion of hot fluid that periodically advances along the top of the pipe. This hot tongue erodes and effectively "swallows" portion of cold and stratified layer frequently, leading to a specific pattern of thermal loading along the pipe's length.

This work provides a comprehensive experimental database that enhances the physical understanding of mixing in dead-legs and serves as a valuable resource for flow topology and parameter influence in dead-legs flow.

MYRRHA (Multi-purpose hYbrid Research Reactor for High-tech Applications) is a flexible fast-spectrum, pool-type research reactor under construction at SCKCEN, cooled by Lead Bismuth Eutectic (LBE). It is identified as the European Technology Pilot Plant for the Lead Cooled Fast Reactor, one of the GenIV reactor concepts.
In the framework of the Euratom ANSELMUS project, the SARDINE experiment was setup at SCKCEN within the COMPLOT (COMPonent LOop Testing) facility for the construction of the MYRRHA Safety Rod (SR). The SARDINE experiment is indeed a proof of concept of the timely insertion of the SR. The SRs act as a backup mechanism for the safe reactor shut down in case of emergency shutdown. The full downward displacement of the absorber bundle assembly should be completed within 1.0 s. The SARDINE conceptual design is supported by a simplified 1D numerical model, mainly based on the interplay of two reciprocal pistons.
In parallel with the experimental facility construction, a numerical CFD model of the SR test section was built at CRS4 in order to: (i) numerically reproduce the SR tests in SARDINE, (ii) verify that the insertion time is acceptable and (iii) validate the CFD simulations against the experimental results.
This paper describes the pre-test numerical activities in support of the experimental campaign. The conceptual design of the experimental setup is based on preliminary one-dimensional calculations developed by SCKCEN for the evaluation of the transient displacement. The impact of the LBE flow on the safety rod displacement is difficult to evaluate in a 1-dimensional model. For this, CRS4 developed a full 3D CFD model of the SR displacement which couples with the 1D model in order to improve its accuracy.
The paper is organized as follows. At first, we describe the one-dimensional model together with the SR operating principle. It is implemented in MATLAB and simulates the insertion movement of the SR as a result of the forces acting its moving components: (i) LBE buoyancy, (ii) components’ weight, (iii) LBE drag, (iv) gas pressure in the pneumatic pistons, (v) spring compression between moving components and (vi) bottom impact hard spring. The focus is put on the pressure gas pistons system which delivers the main driving force. Second, the experimental setup is described. The SR consists in a bundle of 12 absorber rod, connected to the pistons system. A tungsten ballast is placed above the absorber rods bundle, compensating its positive buoyancy. Then, we describe extensively the pre-test numerical CFD simulations performed by means of the STAR-CCM+ code.
The simulation of the SR insertion is based on the overset mesh method. The complex geometry is represented with high fidelity where feasible. A high level of detail could be kept in the geometrical modeling of the pin bundle wheel-guided or the hollow perforated tube enclosing the ballast. The background domain and two volumes enclosing the moving components are represented by separated overlapping regions, each one with its own mesh. Three volume interfaces of few cell layers are created in order to couple the solutions on the three overlapping grids, with no need for re-meshing.
In order to be representative of MYRRHA, the Volume of Fluid (VoF) framework is considered. The SR dynamics requires the inclusion of three separate free surfaces: two located above the SR and one in the bottom buffer tank. Suitable boundary conditions have been determined for the initial steady state simulation. The two-phase flow must first be stabilized at the LBE nominal mass flow rate which serves as initial condition for the successive transient simulation of the SR effective insertion.
The fully explicit two-way coupling obtained from the balance of all the forces acting on the SR is implemented, providing the downward acceleration to the bundle assembly and to the ballast. The CFD model combined with the 1D model predicts the timely safety rod insertion, consistently with the 1D model alone. Both models’ results are discussed and compared in the following. The simulations distinguish between the MYRRHA scenario with high initial gas temperature (220°C) and the COMPLOT scenario with initial gas room temperature (20°C). The transient evolution of the positions, velocities, gas pressures and individual forces acting on the bundle assembly and the ballast is monitored and reported. For the nominal MYRRHA scenario, an effective insertion time well below the target value of 1.0 s is predicted. In the COMPLOT scenario, a slightly longer insertion time is predicted, thus providing a conservative estimation of the SR insertion time in MYRRHA. At last, a degraded configuration is studied. The gas feed is assumed to have stopped and the venting lines are blocked. The SR is passively inserted in about 10 s.
The models are ready for the comparison with the experimental tests and for the post-test validation phase.