This paper focuses on the identification/prediction of the blade exhibiting the largest response in mistuned bladed disks. This information is very important in experimental/testing efforts as it permits the most effective positioning of a few gages to capture the maximum response on the disk. In computational statistical analyses, knowing the highest responding blade is also quite valuable as it may lead to computational savings in the determination of the maximum response. Different strategies are proposed here for the experimental and computational contexts. In the former situation, mistuning is typically unknown but only one or a few disks must be considered. The proposed solution is then to estimate the mistuned blade properties and to rely on this identified bladed disk model to predict the blades that are likely to exhibit the largest responses through exact, full disk solutions. On the contrary, in computational statistical analyses, mistuning is specified but a potentially large number of disks must be analyzed and it is desired to bypass the ensemble of full disk solutions. Accordingly, a novel, computationally very efficient algorithm is proposed for a preliminary estimation of the forced response of mistuned disks from which the blades that are likely to exhibit the largest responses can be predicted. Examples of application on single- and two-degree-of-freedom per blade models and a reduced order model of a blisk demonstrate the reliability of the proposed strategies.

This content is only available via PDF.
You do not currently have access to this content.