Aeroelastic Instabilities due to Unsteady Aerodynamics
One of the grand challenges faced by industry is the accurate prediction of unsteady aerodynamics events, including frequency lock-in and forced response. These aeromechanical incidents occurring in airplane engines and gas turbines can cause high-amplitude blade vibration and potential failure of the engine or turbine. During the last decades, the development of computational fluid dynamics has allowed the design and optimization of complex components while reducing the need for expensive engine testing. However, the validation of frequency lock-in and forced response numerical results with experimental data is very incomplete. Despite tremendous advances in computational capabilities, industry is still looking to validate design tools and guidelines to avoid these potentially costly aeroelastic events early in the design process.
The research efforts presented in this dissertation investigate the aeroelastic phenomena of frequency lock-in and forced response in turbomachinery. First, frequency lock-in is predicted for two structures, namely a two-dimensional cylinder and a single three-dimensional airfoil, and the results are compared to experimental data so that the methods can be extended to more complex structures. For these two simpler structures, a frequency domain harmonic balance code is used to estimate the natural shedding frequency and the corresponding lock-in region. Both the shedding frequencies and the lock-in regions obtained by an enforced motion method agree with experimental data from previous literature and wind tunnel tests. Moreover, the aerodynamic model of the vibrating cylinder is coupled with the structural equations of motion to form a fluid-structure interaction model and to compute the limit-cycle oscillation amplitude of the cylinder. The extent of the lock-in region matches the experimental data very well, yet the peak amplitude is underestimated in the numerical model. We demonstrate that the inclusion of the cylinder second degree of freedom has a significant impact on the cylinder first degree of freedom amplitude. Moreover, it is observed that two harmonics need to be kept in the equations of motion for accurate prediction of the unsteady forces on the cylinder.
The second important topic covered is a comprehensive forced response analysis conducted on a multi-stage axial compressor and compared with the initial data of the largest forced response experimental data set ever obtained in the field. Both a frequency domain and a time domain codes are used. The steady-state and time-averaged aerodynamic performance results compare well with experimental data, although losses are underestimated due to the lack of secondary flow paths and fillets in the model. The use of mixing planes in the steady simulations underpredicts the wakes by neglecting the important interactions between rows. Therefore, for similar cases with significant flow separation, the use of a decoupled method for forced response predictions cannot yield accurate results. A full multi-row transient analysis must be conducted for accurate prediction of the wakes and surface unsteady pressures. Finally, for the first time, predicted mistuned blade amplitudes are compared to mistuned experimental data. The downstream stator is found to be necessary for the accurate prediction of the modal forces and vibration amplitudes. The mistuned rotor is shown to be extremely sensitive to perturbations in blade frequency mistuning, aerodynamic asymmetry, and excitation traveling wave content. Since this dissertation presents the initial results of a five-year research program, more research will be conducted on this compressor to draw guidelines that can be used by aeromechanical engineers to safely avoid forced response events in the design of jet engines and gas turbines.
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 3.0 United States License.
Rights for Collection: Duke Dissertations
Works are deposited here by their authors, and represent their research and opinions, not that of Duke University. Some materials and descriptions may include offensive content. More info