Mistuning Identification of Blisks Using Mass Detuning and Influence Coefficients

Abstract

Integrally bladed disks, also known as blisks, are important components in the operation of gas turbines. To study the dynamics of these structures, several computational techniques approximate blisks as tuned systems, where all blades have the same characteristics. In reality, manufactured blades have small variations from their intended design, thus creating so-called mistuned systems. Previous work has shown that such small variations can cause high response amplifications that can result in blisk structural failure. Therefore, it is essential to identify mistuning to ensure the safe operation of these systems. Several methods utilize cantilever blade frequencies in their procedure to model mistuning and calculate blisk vibration response. A previously developed mistuning identification method uses mass detuning to isolate the cantilever blade frequencies and a set of influence coefficients to quantify the blade-to-blade couplings. However, the mistuning is calculated with respect to an arbitrary mistuning value that later needs to be identified through a trial-and-error procedure that can affect the accuracy of the final mistuning values. This paper focuses on updating the method to eliminate this additive error that cannot precisely be determined. Furthermore, the updated procedure is applied to an as-manufactured blisk using experimental ping test results. An additional analysis is performed on the effect of adding multiple sets of influence coefficients to capture the blade-to-blade couplings more accurately.