Reduced order modeling and mistuning identification method for rotating bladed disks under varying speeds

As a critical component in turbomachinery, such as aero-engines, bladed disks frequently experience mistuning due to various factors, leading to localized vibrations and increased risk of high-cycle fatigue. To enable online mistuning identification and dynamic response prediction of rotating bladed disks, this paper proposes a variable-speed reduced order model (VSROM) that accounts for varying rotational speeds and a response-based mistuning identification method. By parameterizing the stiffness matrix of the bladed disk as a polynomial and assuming the tuned system's mode shapes are minimally affected by speed, the VSROM can be developed using the Component Mode Mistuning method. Additionally, leveraging the VSROM, a response-based mistuning identification method is proposed, capable of identifying mistuning at any speed. The effectiveness and accuracy of the VSROM and mistuning identification method are validated through numerical simulations. The dynamic response of the mistuned system is predicted using the VSROM, and the results are compared with those obtained from the existing reduced order model that accounts for varying speeds, as well as from the full-order finite element model. The results demonstrate that the proposed VSROM offers superior prediction accuracy and computational efficiency. Moreover, mistuning identification at different speeds shows good agreement with the actual mistuning values. The proposed VSROM and mistuning identification method hold significant potential for online vibration monitoring of rotating bladed disks.