Conference Agenda

The interaction between cooling fluid and solid structures (rods, tubes) in nuclear power plants leads to flow-induced vibrations (FIV). These may cause material fatigue, fretting wear, and in worst case, loss of component integrity. The consequence of this might be high standstill costs due to longer or unplanned outages or safety issues like Steam Generator Tube Rupture accidents. Within the European GO-VIKING (Gathering expertise On Vibration ImpaKt In Nuclear power Generation) project, experimental and numerical efforts are performed to improve the understanding and analysis of FIV in nuclear power plants, as well as to develop and validate advanced numerical approaches for their prediction. Within the fifth work package of the project, activities to analyze the tube vibration behavior in the CEA’s AMOVI experiment were carried out at GRS. AMOVI deals with FIV occurring in tube configurations, exposed to a cross-flow. In this work, the FIV are investigated with a fluid-structure interaction (FSI) approach, based on a structural reduced-order model. Such models are of particular interest due to the excessive computational time necessary for the FIV evaluations. The reduced-order model for the structural domain MOR was coupled to ANSYS CFX for the FSI calculations presented in this paper. This FSI approach was validated against AMOVI data for a single flexible tube positioned in the center of a bundle of rigid tubes.

Fluid-induced vibrations (FIV) are a major cause of component failure in nuclear reactors and are of significant concern when extending operating reactor lifespans. The Go-Viking project aims to ensure that the licensed operating lifetime of aging nuclear reactors in Europe can be safely extended by improving understanding of FIV phenomena within steam generator tube bundles. To achieve this, the GOKSTAD experimental tube bundle has been created to collect FIV data for an inline cross-flow tube bundle operating at a higher Reynolds number than previously documented in the literature. In this presented work, we present results from high-resolution direct numerical simulations (DNS) of the GOKSTAD bundle for comparison with experimental data. Due to the computational cost of DNS, a reduced domain of the GOKSTAD bundle is used in these simulations, consisting of three rows with seven columns (five regular columns and two half-tube columns). The mass flow rate within the tube bundle is 15 m³/s, and the pitch-to-diameter ratio within the bundle is 1.44. The DNS is done using the Department of Energy code, NekRS, while the structural responses are simulated using the Multiphysics Object-Oriented Simulation Environment (MOOSE). Results include velocity fields, pressure fields and displacement data of the center tube, allowing direct comparison with measurements collected within the GOKSTAD bundle. The DNS FIV model generated from these results will support creation of fast-running FIV tools, including reduced-order models.

Key Nuclear Steam Supply System (NSSS) components, such as fuel rods and steam generator tubes, are highly susceptible to Flow-Induced Vibrations (FIV) as a result of the turbulent coolant flow. This can cause known failures of these components, such as Grid-To-Rod-Fretting (GTRF) wear and Steam-Generator Tube Rupture (SGTR), possibly leading to costly reactor outages. Historically, analytical and semi-empirical approaches were used to assess the impact of FIV on the components’ structural integrity. However, these are generally only able to give an order of magnitude estimate of the structure’s displacement. With the increase in computational power though, Fluid-Structure Interaction (FSI) simulations, two-way coupling detailed Computational Fluid Dynamics (CFD) and Computational Structural Mechanics (CSM) codes, are being used more and more. Such FSI simulations are able to give increasingly better predictions, matching reference experimental data quite well, in particular in terms of vibration frequency and displacement amplitude. These simulations generally only consider cylinders undergoing relatively small displacements, avoiding contact. To capture GTRF or SGTR resulting from FIV, generally larger displacements are needed, along with contact between neighboring cylinders or between a cylinder and surrounding fixed structural components. The current work shows results of an initial investigation of performing FIV simulations involving structural contact. It considers a cylinder placed inside a channel and subjected to turbulent cross-flow. Different numerical and modeling techniques have been used to try to successfully resolve the large displacements and structural contact. These are presented, along with results of a first FIV simulation involving contact between the cylinder and the channel walls.

Understanding the behavior of flow-induced vibrations (FIV) is crucial for maintaining the safety of steam generators in nuclear power plants. If left unaddressed, vibrations can lead to tube wear, fatigue, and even failure, which can have profound safety consequences. In U-tube designs where two-phase cross-flow dominates, vibration-related issues are further exacerbated. Fluid-elastic instability is the primary mechanism underlying flow-induced vibration that may damage steam generator tubes. Although fluid-elastic instability in single-phase cross-flow has been extensively studied, its behavior in two-phase flows is less understood. The Horizon Europe project GO-VIKING addresses these challenges through experimental facilities designed to study two-phase cross-flow-induced vibrations. These setups focus on the fluid-structure interaction (FSI) between tube bundles and air-water cross-flows.

This paper presents numerical simulations of FIV in two-phase cross-flows. Simulations cover two-phase flows over single tubes and tube bundles. On the fluid side, we use a two-fluid model coupled with a population balance model to account for bubble poly-dispersity, coalescence, and break-up. The structure motion is modeled using a six-degree-of-freedom rigid body motion solver. The numerical results are validated against experimental data on bubble size distributions, void fractions, fluid-structure forces, and displacement spectra. The findings of this work advance the understanding of two-phase FIV, providing insights critical to the safe and reliable performance of steam generators.

In this work, the vibration characteristics of a propeller rotating in lead-bismuth eutectic (LBE) are studied using fluid-structure interaction (FSI) simulations, which are developed in parallel with an experiment performed at the Belgian Nuclear Research Center (SCK CEN). Both the simulations and the experiment are part of a larger campaign to develop a methodology for characterizing the vibrations in primary pumps of nuclear reactors using heavy liquid metal coolants. These coolants are of interest due to their use in Generation IV nuclear facilities such as MYRRHA , using LBE as a coolant, and LFR, using lead. The high density of this liquid can significantly alter the vibration characteristics compared to when used in air and water, and introduce mode coupling, a phenomenon that is not yet sufficiently understood in the context of heavy liquid metals. The simulations allow for an extensive analysis of the different vibration modes. The investigated propeller consists of three symmetrical blades and is operated at different rotational speeds. First, the structural eigenmodes are calculated in vacuum, using the finite element method (FEM). Afterwards, the fluid and propeller are combined in a two-way coupled FSI simulation. For each mode a particular excitation force is applied to the structure to facilitate the extraction of vibration characteristics by analyzing the free response of the system. The result is the eigenfrequency and damping ratio of that mode in LBE. The results show that this methodology allows for an accurate prediction of the measured vibration response in the test setup.

To improve the long-term operation of NPP, dedicated tools are needed to understand and predict the interaction between cooling fluid and solid structures that may lead to flow-induced vibrations. The paper focuses on a validation test case representative of a steam generator configuration.

In the GO-VIKING project, supported by Horizon Europe research and innovation funding, a new water loop named GOKSTAD has been designed and constructed at the von Karman Institute to characterize a single-phase flow field and study the fluid-structure interaction inside a 7*7 square lattice in cross-flow configuration. The facility is designed to operate up to a remarkable gap Reynolds number of 90,000; a significant leap beyond what is currently available in literature.

The paper will first present the facility's flow field characterization based on PIV measurements at the inlet of the test section, between the cylindrical rods of the lattice, and at the outlet of the bundle. The second part will present the mechanical design and characterization of the moving rigid cylindrical tube developed for the Fluid-Structure-Interaction study. The configuration studied consists of two vibrating tubes inline or side by side inside the square lattice, with a vibration of roughly maximum of 10% of the cylinder diameter.

The results presented are precious data for the numerical team in charge of fluid-structure interaction studies, providing high-resolution boundary conditions and flow field data. The final objective is to have medium-resolution numerical tools to assess structural vibrations in Steam Generators under single-phase cross-flow conditions.