Context: The DyVA industrial chair, co-financed by the ANR (French National Research Agency) and the SAFRAN Group, aims to meet the environmental challenges facing the aeronautics sector, involving a drastic reduction in CO2 emissions over the medium and long term, with a target of low-carbon by 2035 and carbon neutrality by 2050. Technological programs such as RISE, launched by SAFRAN, will need to accurately analyse new operating points and their impact on engine dynamics and service life. Indeed, this project addresses a large number of technologies that break with conventional architectures, such as open-fan, open-OGV and fast turbine and booster. 

DyVA aims to develop advanced numerical tools capable of meeting the challenge of vibration prediction for new aeronautical engines. The planned developments will focus on the simulation and modelling of non-linear dynamic behaviour and uncertainties, in order to provide in-depth knowledge of the underlying dynamics of the system and master its physics, simulation and all the different possible operating points. These results will be correlated with experimental tests, providing measurements that are often lacking in the literature, but are essential for a good understanding of the physics.

Your tasks: Dynamic response prediction validation by tests involves the best possible command of the load applied to it. Under flow experiments require considerable investments such as wind tunnels, and can include potentially destructive unstable aeroelastic forcing conditions such as flutter. In DyVA, many experiments will be performed under vacuum to focus on mechanical behaviour: an innovative approach will consist in replacing aerodynamic loads by an electromechanical load applied using piezoelectric patches, so as to mimic flutter, wakes, vortex ingestion or even crosswind thus providing a safe, versatile and economical tool for testing new technological solutions. Such excitation strategies exist to generate forced vibrations on cyclic symmetry structures at rest [1]. Concerning the use of unstable diverging conditions for mechanical characterization, very few attempts have been made, and only on very simple static structures [2]. This PhD aims to go further in excitation realism, especially when unstable conditions arise, emphasis being placed here on mimicking flutter. 

Fluid-structure interaction will be simulated through the constitutive law coupling the behaviour of piezoelectric patches and structures. The objective is to propose a hardware-friendly excitation device to test the vibratory forced response of industrial structures in other DyVA workpackages. Indeed, the benefit of the strategy proposed here is all the more effective when testing damping devices that require centrifugation to work, such as dovetail blade attachments, friction rings or under-platform dampers for which some studies have already been performed experimentally at LTDS [3]. 

A  multi-physics numerical  activity  is required  based  on the  formulation  of a  control-loop  that mimics,  through piezoelectric excitation, aerodynamic loadings focused on flutter. This process will exploit simplified aerodynamic effect using    inter-blade coupling terms or machine-learning approach based on existing CFD aerodynamic data and vibratory experimental data, including piloting and control issues. Excitation laws will be built up as a "hardware in the loop" strategy, tuned to reach the desired level of fidelity.