Master thesis and internship[BR]- Master's thesis : Analysis and simulation of aeroelastic effects in turbomachinery[BR]- Internship

Abstract

This thesis investigates the structural dynamics of turbomachinery components, with a focus on the application of reduced order modeling techniques to enhance computational efficiency. The Craig-Bampton substructure coupling method is employed to reduce the computational time required for the dynamic analysis of a bladed-disk model while maintaining accuracy. The study is conducted using two distinct strategies: one involving a change of variables applied before the reduction method, and the other without such a transformation. These strategies are compared in terms of convergence and accuracy across various scenarios, including a realistic bladed-disk sector model.

In addition to the linear analysis, the thesis extends its scope to incorporate nonlinearities, recognizing the significant impact that nonlinear elements can have on the dynamic response of real-world systems. Nonlinear polynomial stiffness and Coulomb friction are introduced into the reduced models to capture these effects. The nonlinear analysis is performed using harmonic balance continuation, and the results are compared across both reduction strategies to evaluate their effectiveness in accurately representing the nonlinear dynamics.

The findings demonstrate that while both strategies yield consistent results in the linear analysis, the nonlinear dynamics introduce complexities that are captured differently by each approach. The study concludes that the Change of Variables (COV) method offers a slight computational advantage and maintains accuracy, particularly in scenarios involving nonlinear interactions.

This work not only validates the use of the Craig-Bampton reduction method for efficient dynamic analysis but also provides insights into the optimal strategy for incorporating nonlinearities in reduced-order models. The results are significant for advancing the modeling and analysis of turbomachinery components, where both linear and nonlinear dynamic behaviors play a critical role in performance and safety.