Simulation of Fluid-Structure Interaction of Crossflow Through a Tube Bundle and Experimental Validation

Abstract

Fluid-structure interactions are complex, multi-physics phenomena of consequence for many fluid-flow domains. Modern multi-physics codes are becoming capable of simulating with great accuracy the interaction between fluid and structure dynamics. While fluid-structure interactions can occur in many forms, flow-induced vibrations are of particular interest. Such vibrations can result in the fatigue and even failure of a vibrating geometry. The prediction and minimization of flow induced vibrations are of particular importance for heat exchangers, which commonly contain bundles of tubes experiencing high-velocity crossflow. The present study simulates the fluid-structure interaction for flexibly mounted tube bundles undergoing crossflow and compares the results with experiment. The simulation code consists of a spectral-element fluid solver directly coupled with a finite-element solid mechanics solver. The fluid solver locally adapts the fluid mesh to accommodate the moving solids. In order to minimize computational expense, low Reynolds number flows are considered, allowing for a thin, pseudo 2-D domain. The flow remains laminar for the majority of the domain, with local areas of turbulence. The pins are connected to springs that supply a restorative force equivalent to the flexible mounts of the corresponding experiment. Fluid-only simulations are performed for flow spanning low to moderate velocities and compared visually with experimental flow visualizations. Coupled fluid-structure interactions are simulated with low velocity and vibration amplitudes. The measured vibration amplitudes of the simulation agree well with those of the experiment.

Demonstration of Fully Coupled Calculations in Helical Steam Generator: Toward Predictions of Fluid Elastic Instability

“…Here, we employ the latter approach due to its higher flexibility of combining nearly any kind of single-field formulation and straight-forwardness when coupling independent solvers such as Nek5000 (fluid) and Diablo (structure). The strategy of the coupled solver is discussed in detail in [28,29] along with some validation results. In particular an acceleration strategy employing fictitious mass damping is discussed.…”

“…An advanced numerical simulation capability for modeling such phenomena will help improve the analysis and evaluation of different design variants in terms of vibrations and heat transfer performance, thus complementing expensive experimental tests and reducing their cost, while developing a better 2 understanding of the physics behind FIV. Work performed in the previous year lead to the development of such a tool in the multi-physics suite SHARP, partially validated on classic fluid-structure interactions (FSI) benchmark data [28,29].…”