Session: VIB-01/MSNDC-08-03

Paper Number: 148154

148154 - A Generalized Mistuning Modeling Technique for Bladed Disks 

Bladed disks always have variations between blades, known as mistuning, which is a focus of a significant amount of research within the turbomachinery industry. When mistuning is present in nominally cyclic structures, the forced response of the system can show very localized amplification of blade responses when compared to its tuned system. Due to the random nature of these variations often a statistical analysis is conducted to understand a bladed disk design’s sensitivity to mistuning. For high dimensional industrial models this tends to be an intractable problem unless the model is reduced. To this end, several types of reduced order models (ROMs) have been developed that enable the projection of the mistuning in the reduced space to greatly reduce computational expense and make statistical analysis of very complex blade designs possible. These methods include techniques for reducing systems with a variety of types including modulus variations, geometric variations, and contact variations of various levels.

This work aims to create an efficient generalized mistuning model (GMM) that enables the incorporation a number of forms of mistuning at any level into the blade or disk model. The GMM method has many of the same benefits of previous mistuning models by being very compact and only requiring sector level models and calculations, which makes it useful for high dimensional industrial models. The GMM method utilizes the Craig-Bampton component mode synthesis to decouple the blade and disk sector. These models can then be projected onto a reduced modal space with any number of forms of mistuning (i.e., small, large, mass, frequency, geometric) efficiently applied to the blade and/or disk models. Note that small frequency mistuning will only require a single sector model to create many realizations, however large geometric mistuning will require different blade models to be analyzed to project these blades into the reduced space.  The combination of the blade and disk reduction occurs through the projection of the interface onto special disk-blade interface modes, and then enforcing compatibility. A few additional conditioning steps are applied to ensure a stable and invertible transformation matrix.  A numerical validation of the developed method is performed for a few types of mistuning to show the power of the GMM for effectively handling multiple types of mistuning.

Presenting Author: Kiran D'souza Ohio State University

Presenting Author Biography: Dr. Kiran D’Souza became an Associate Professor at The Ohio State University (OSU) in 2020, and serves as the Associate Director of the Gas Turbine Laboratory (GTL) and Director of the Nonlinear Dynamics and Vibration Laboratory (NDVL). He obtained a PhD in Mechanical Engineering from the University of Michigan (UM) in 2009 and then worked as a postdoctoral researcher and Research Scientist at UM before joining OSU in 2014.

Authors:

Kiran D'souza Ohio State University
Troy Krizak Ohio State University