Modelli aeroelastici computazionali per compressori e fan ad alta velocità

International regulatory institutions have set ambitious targets to drive the development of commercial aviation towards a sustainable growth and a net zero carbon emission by 2050. Some of the proposed solutions to increase the sustainability of air transport include the development of ultra-high bypass ratio engines, fuselage-embedded propulsors, a general reduction of weight and an increase in rotor velocity, leading to a higher blade loading. These design trends eventually bring about a reduced mass ratio (that is the ratio between blade mass and that of surrounding flow) and increased disturbances when boundary-layer ingesting solutions are considered, leading to a potentially higher aeroelastic sensitivity. Flutter in turbomachinery is a dangerous aeroelastic phenomenon to cope with, potentially leading to catastrophic failures, hence an accurate prediction of its insurgence in typical operative conditions is crucial. Experimental tests are expensive and difficult to conduct, and running unstable configurations can damage the machine: for this reason, computational flutter prediction techniques are fundamental. However, the inherent nonlinear nature of transonic flows and the necessity to model complex fluid-structure interactions pose relevant modelling difficulties and lead to the development of an array of flutter computational techniques, with a largely variable fidelity, as well as several reduced order models to address the preliminary assessment of flutter stability in new compressor geometries. This work aims at the development of a reduced order model (ROM) for flutter prediction in transonic compressors and fans. The ROM is based on the aerodynamic influence coeffcients (AIC) technique, which allows to model the aerodynamic loads induced by the deflection of one blade on the neighbouring ones using an unsteady simulation and system identification technique, exploiting the linearity hypothesis. With this formulation, both uncoupled and coupled aeroelastic modelling can be addressed, and multiple inter-blade phase angles (IBPA) can be simulated without the need of further RANS computations. In the first part, the main computational techniques for flutter simulation in turbomachinery are reviewed, and the principal model order reduction techniques are introduced. In the second part, the developed ROM is applied to the TU-Darmstadt Open Test transonic rotor. Two resonance conditions are investigated, where the experimentally measured aerodynamic damping exhibits its minimum values. Both an uncoupled and a coupled formulation are adopted to assess the aeroelastic stability at the two resonances for a range of mass flows. The uncoupled model results are compared to classical Fourier transformation and harmonic balance methods, while the closed-loop coupled model are compared to experimental data on damping ratio.