Winkelmann, U. (2021)
Abstract
The design of engineering structures resilient to aeroelastic Fluid-Structure Interaction (FSI) remains as one of the fundamental challenges in the field of wind engineering. Vortex-Induced Vibrations (VIV) at lock-in, still represent a not entirely understood FSI-born phenomenon with a rather high prediction uncertainty regarding the amplitudes of the arising structural oscillations. In order to relieve existing conservative VIV-response estimations, which compensate the lack of understanding, a systematic assessment of numerical predictions by means of validation against experimental data is crucial. In this context, the work presented within this thesis applies a hybrid strategy. Namely, the concept utilizes Computational Fluid Dynamics (CFD) simulations and experimental observations to assess the simulations‘ accuracy and their applicability for wind engineering purposes. Then, the CFD simulations are utilized to gain additional insight into the nature of VIV beyond the relevant experimental data. More precisely, the research focuses on aeroelastic VIV in the lock-in region of turbulent flows in subcritial and transcritical Reynolds number (Re) regimes for rigid, infinite circular cylinder sections. 2D Unsteady Reynolds-Averaged Navier-Stokes (URANS) simulations with the k -omega SST and 3D Large Eddy Simulations (LESs) with the Smagorinsky turbulence model are conducted for a stationary, forced vibrating and free vibrating body. Despite the discrepancy of a few aerodynamic and aeroelastic parameters, the obtained results show that URANS simulations delivered fairly good qualitative predictions compared to the results from wind tunnel data. The results of the LESs at subcritical Re are in even better agreement with experimental data where most parameters comply quantitatively with wind tunnel results. Hence, LESs are deemed appropriate as a complementary tool for wind engineering applications of VIV around rigid circular cylinders for the structural oscillation amplitude prediction due to aeroelastic VIV. The thesis also discusses revealed inconsistencies and parameters‘ sensitivity of various quantities. Numerical simulations are indeed an important tool for enhanced insight into the complex flow phenomena beyond the limitations of performed experiments. Based on the presented
results and appraisal, it is clearly illustrated that these strategies enable a more inclusive approach for assessing VIV. This work presents a step towards building such an effective methodology aimed at unraveling existing wind engineering conundrums.