Single stage and multi stage effects of mistuning on the response of blades

The focus of the present investigation is on three specific problems relating to the presence of small blade-to-blade variations in their geometry and/or material properties, collectively referred to as mistuning. The first part of this dissertation investigated the possibility to damp entire bladed disks using linear dampers on only a fraction of the blades. This problem was first addressed without the presence of random mistuning and an optimization effort was undertaken to select the locations of an increasing number of dampers to reduce as much as possible the maximum amplitude of blade response on the disk. This process led in particular to a significant, steady reduction in the amplification factor with increasing number of dampers. More specifically, it was found that a single damper does not lead to any noticeable reduction in forced response because of localization issues and an appropriate alignment of the traveling mode shapes. However, two and three dampers did lead to very significant reduction in forced response. Unfortunately, the presence of random mistuning negated much of the benefits of the introduction of a subset number of dampers.;The second part of this investigation focused on the effects of mistuning at the blade-disk interface that may originate from the seating of the blade in the disk (for inserted blades) or from variations of geometry in the fillet area (for blisks). To this end a blade-interface-disk mean model was first developed that relies on both the Craig-Bampton approach and a local modeling of the interface. Then, both model and data uncertainties were introduced in this model using the nonparametric stochastic modeling approach. An example of application was next considered to exemplify these concepts and provide a comparison with the more classical blade mistuning effects.