The analysis of the structural dynamic properties of nominally cyclic structures, such as bladed discs, is often done under the assumption, that the whole structure is comprised of perfectly identical (tuned) substructures or sections. But it has been shown, that the dynamic response of a bladed disk is highly sensitive to small variations (random mistuning) in geometry and material properties among the sectors which are caused e.g. by manufacturing tolerances, material defects and uneven wear. These differences, called mistuning, can have a large impact on the dynamic behavior, leading to a response which is qualitatively different from that of an ideal tuned structure.
In particular, the phenomena of peak splitting (splitting of double eigenvalues) and mode localization (containment of vibration energy in specific sections of the cyclic structure) can lead to a dramatically increased amplitude of vibration (displacement amplification, see figure 1). Some high sensitive parameters affecting the modified, mistuned behavior of the blisk are the standard deviation of the mistuning strength, in combination with it´s distribution, the structural coupling between the disc sections and the modal density of it´s natural frequencies.
Here, at the institute of Aeroelasticity, a test rig will be designed and manufactured to simulate engine order excitation in stationary (non-rotating) bladed disks for the purpose of identifying free and forced response mode localization and displacement amplification due to mistuning. The experimental results of the tuned and some mistuned blisks will also provide a database for updating the corresponding numerical models and to verify the accuracy of some reduced order models (ROM) for probabilistic investigations.