Recently, an efficient approach for modeling the vibration of multistage rotors was developed by the authors [1, 2]. This reduced-order modeling technique employs component mode synthesis, with each stage (bladed disk) treated as a separate component. In addition, the component mode mistuning (CMM) projection technique was extended to multistage systems. In the CMM method, individual blade mistuning is transformed from a basis of cantilevered blade modes to the basis of tuned system modes used for the reduced-order model. In this paper, the component-based modeling framework developed for mistuned multistage turbine engine rotors is utilized for system identification. First, the identification of multistage mode types is considered. Strain energy ratios are used to identify which system modes are confined to mostly one stage and which modes show strong coupling among multiple stages. Simple approximations for these ratios are derived based on data from the component-level free response analysis that are performed during the model construction process. The component-level results are also utilized to identify the dominant nodal diameter number for each multistage mode, even though the multistage system does not possess cyclic symmetry because the stages have different numbers of blades. Second, the modes are further classified as to how much the blades participate in the response relative to the disk for each stage. As a systematic identification procedure, this is applicable to single-stage models as well. For multistage systems, this is used to determine operating conditions where coupled response among blades on adjacent stages is most likely to occur. Third, the application of mistuning identification techniques to multistage systems is considered. It is found that the proposed modal classification methods allow the determination of conditions under which deviations in individual blade properties may be observed indirectly from measurements of the disks and spacer.

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