A turbine’s structural response is difficult to predict because turbine blade systems tend to have modes with closely spaced frequencies. When the natural frequencies are closely spaced, small changes in the structure, e.g., blade frequency variations, can cause large changes in the mode shapes and the turbine’s vibratory response. As a result, the amplitudes of blades can vary significantly from one blade to the next and from one engine to the next. This causes a large amount of uncertainty when predicting the fatigue life of these components.

    Professor Griffin will explain the underlying cause of the mistuning problem and recent research in developing reduced order models that capture its physical behavior. The reduced order models have the structural fidelity of a finite element analysis of the full bladed disk and the computational efficiency of mass-spring models. Because the methods are highly efficient they can be used in Monte Carlo simulations to quickly determine the response of thousands of randomly mistuned bladed disks. The resulting data can be used to estimate the statistical distribution of blade amplitudes for a fleet of aircraft and assess the likelihood of blades failing from high cycle fatigue.

    The talk will focus on a new, fundamental model of mistuning recently developed at Carnegie Mellon University. It is fundamental because of its simplicity and the fact that it identifies the critical parameters required to represent mistuning. It will be shown that it can be used for system identification as well as prediction.