Investigations on lock-in vortex-induced vibration of an airfoil at a high angle of attack based on detached eddy simulation

Abstract

Large-scale modern wind turbines at standstill are prone to vortex-induced vibration (VIV). In this study, coupled fluid–solid dynamics of the wind turbine airfoil at a 90° attack angle are performed using the detached eddy simulation. The fully developed vibration responses with different structural dampings are explored in detail. The frequency lock-in regime is determined, and the corresponding phase differences between the lift and displacement are presented with the Lissajous curve. The dominant surface pressure mode and wake flow exhibit significant three-dimensional flow characteristics in unlock-in conditions, while a strong spanwise correlation in lock-in conditions is detected. The pressure fluctuation on the suction side in the lock-in state is observed to be more significant than in the unlock-in state. The effect of the distributed airfoil surface pressure on VIV is evaluated by considering the contribution value and the cyclic aerodynamic work density. With the decrease in structural damping, the aerodynamic work near the leading edge gets enhanced and the negative work region is reduced, leading to a higher amplitude of VIV. The beat vibration and hysteresis behavior at the critical reduced velocity are also analyzed in both the time domain and frequency domain.

High-fidelity simulations of airfoil vortex-induced vibrations: from 2D to blade-like aspect ratios

“…The results obtained are shown in Table 2, along with some values obtained for similar cases from the existing literature, all with the same airfoil. The Strouhal number for the 2D case is 0.126, which is in line with previously reported values for the same airfoil and with URANS, namely 0.125 by [9], 0.121 [13] by and 0.119 by [10]. The Strouhal number for the 3D1c case is 0.132, which is somewhat far from previously reported values for the same airfoil, namely 0.153 by [9] with DDES, and also 0.153 by [10] with DES.…”

“…The Strouhal number for the 2D case is 0.126, which is in line with previously reported values for the same airfoil and with URANS, namely 0.125 by [9], 0.121 [13] by and 0.119 by [10]. The Strouhal number for the 3D1c case is 0.132, which is somewhat far from previously reported values for the same airfoil, namely 0.153 by [9] with DDES, and also 0.153 by [10] with DES. The authors have found that for the 3D1c case there is a large uncertainty regarding the definition of the 5 Strouhal number, due to the presence of intermittent flow states with different vortex shedding frequencies.…”

“…Thus, the total number of elements in both 3D meshes is about 12.6 million, the 3D1c mesh being 21 times denser in the spanwise direction than the 3D21c mesh, as can be observed by comparing the center and right plots of Figure 1. The level of spanwise mesh density of the 3D1c case is similar to that used in simulations of airfoil sections with similar aspect ratios [9,10], whereas the spanwise density of the 3D21c case is similar to that from simulations of full-blade geometries [4,5,11]. No thorough mesh convergence analyses have been made for the 3D meshes, due to the large computational cost of the simulations.…”

“…The dimensionless time step used in all simulations is ∆ ∆ t = ∆tU ∞ /c = 0.01, with U ∞ being the inflow velocity, which depends on the simulation. The dynamic system characteristics and the time step are comparable to those employed in previous similar simulations [3,9,10,13], and are considered to be representative of a wind turbine blade section.…”