Mistuning in bladed disk assemblies

Abstract

In the framework of Safran Aero Boosters (SAB) R&D mistuning studies, this project aims to determine and evaluate the random mistuning impact of the SAB Engine 3, which also involves studying the influence of the assumptions and uncertainties of the analyses. Mistuning response is computed with a mistuning methodology, which involves Reduced Order Modelling and Component Mode Synthesis for later modal analysis calculation of the rotor, which is a bladed disk. Mistuning random character is modelled by statistical approaches (Weibull and Monte Carlo). Neither aerodynamic damping, nor aerodynamic excitations are known for the Engine 3 modes to be studied. That is why point excitation and estimated aerodynamic damping values have been used. To validate this modelling approach, these 2 terms have been investigated, concluding that: (i) for a given mode, aerodynamic excitation can be replaced by harmonic point excitation with negligibe differences in the results, and (ii) aerodynamic damping small variation does not affect significantly the results, but large variations result in high differences in the amplification factors. Additionally, flutter region studies have been unsuccessfully performed, since AEROMECA (mistuning software) cannot simulate flutter. Then, two modes of interest (A WN, A XN) of Engine 3 have been identified and studied. The mode with highest amplification shows a relatively high amplification with respect to previously studied SAB values. Additionally, its maximum amplification mistuning standard deviation is slightly lower than the values found in other engines. Eventually, since Engine 3 is the 1st studied large engine, it might be suggested that the engine size plays an important role in engines mistuning sensitivity: higher amplification factors for larger engines. Moreover, an statistical study of the normal standard deviation in the curve, has determined an amplification factor close to the maximum one. An Engine 3 veering region mode (B YN) has been studied. This implies assuming several factors (aerodynamic damping, harmonic point excitation shape...). Results show a huge amplification response, but still close to theoretical expectations (4.6% difference). Curve shows a completely unusual trend, never seen in any engine before. Considering the high degree of uncertainties of the analysis, the results are not considered very reliable. In order to identify the unexpected results causes, studies have been performed assessing the influence of aerodynamic damping and harmonic point excitation shape. It has been shown that both can potentially perturbe significantly the results accuracy. This concludes that mistuning studies (particularly those of veering regions) shall be studied with exact parameters and terms.