Session: 28-02 Mistuning: Simulations & Experiments

Paper Number: 102903

102903 - Improvement and Validation of an Autonomous Experimental System to Identify Blisk Mistuning 

This paper presents an autonomous mistuning identification method for single and multiple stages BLISKs. The proposed methodology relies on the ICMM identification methodology. The required inputs are obtained thanks to measurements of the system mode shapes and natural frequencies carried out with a complete test bench. The BLISK is excited by an electrodynamic shaker. A piezoelectric shaker can be added to extend the frequency range and excite higher frequency modes. An infrared Laser Doppler Vibrometer attached to a robotic arm is used to perform the measurements on the blades. The use of a robotic arm enables the reduction of human interventions and significantly limits the positioning error which increases the accuracy of the identification. The proposed setup thus permits to increase the number of measurements and to perform 3D modal analysis.

In this work, the complete methodology is firstly validated considering an industrial BLISK. The focus is then put on the optimization of the input parameters choice. An input data sensitivity study is carried out with the objective of automatization and improved robustness of the process. Especially, the way of selecting the experimental modes and natural frequencies for the mistuning identification procedure is investigated. The improvement of the identification considering 3D modal analysis made possible thanks to the proposed test bench is also explored. This opportunity makes the test bench suitable for next generation of engines which are composed of one-piece multi-stages compressor, the Bladed druM (BluM). Indeed, the overlay of the blade could reduce the accessibility of vertical measurement.

Presenting Author: Sophie Denis V2i

Presenting Author Biography: Sophie Denis graduated from the University of Liege with a Master of Science in Aerospace Engineering in 2021 and is now working in the R&D department of V2i.

Authors:

Sophie Denis V2i
Sébastien Hoffait V2i
Jérôme Ligot V2i
Jean de Cazenove Cenaero
Loïc Salles ULiège