Sensitivity Tests of Medium-Resolution Simulations of Tube Bundles Subjected to Turbulent Cross-Flow

Abstract For efficient heat transfer, tubes of a nuclear power plant steam generator are subjected to turbulent cross-flow. This cross-flow exerts fluctuating forces on the tubes, giving rise to oscillations of the tubes. These oscillations on its turn can result in increased material fatigue and potential damage. Therefore, during the design phase, predictions are made of these so-called flow-induced vibrations (FIV) of the tubes. With the increase in computing power, numerical Fluid-Structure Interaction (FSI) simulations are more frequently used for these predictions. For industrial acceptance and use, medium-to-low resolution tools are needed that can both provide the desired accuracy as well as be operationally practical. However, such tools have given mixed results in the past, and hence more research is needed to better assess the applicability of such tools at simulating FIV of bundles subjected to cross-flow. As part of such efforts, NRG PALLAS is partaking in two different benchmarks of bundles in cross-flow. The first is part of the in-kind VIKING project, and focusses on the VISCACHE 1 experiment performed by EDF in France. The second benchmark centers around GOKSTAD, a new state-of-the-art test facility designed, built and operated by VKI in Belgium as part of the Horizon Europe GO-VIKING project. The current paper presents flow-only and FSI results obtained by NRG PALLAS for these benchmarks, focusing on sensitivity tests to try to gain a better understanding of where seen differences in predictions by the various models come from.