Forced Response System Identification of Full Aero-Engine Rotordynamic Systems

Abstract

A new class of rotordynamic challenges have surfaced in advanced turbofan engine designs where there are opportunities for dynamical coupling between rotor shafts and support structures, requiring entire system-level assessment. Addressing such vibration challenges relies on knowledge of damping levels in the system, but determining the rotordynamic damping in a full aero-engine remains challenging.

This thesis presents a first-of-its-kind forced-response system identification approach to measuring rotordynamic damping of shaft modes in a full gas turbine aero-engine. A reduced-order modeling framework that captures the full-engine dynamics by incorporating the coupling between rotor shafts and support static structure was developed for rigorous design and assessment of the experiment. A statistical analysis involving virtual simulation of the experiment design demonstrates that the experiment is capable of measuring rotordynamic damping within 15% for most shaft modes throughout the operating range of the Pratt & Whitney 615 turbofan engine.

Virtual forced-response system identification experiments demonstrate robustness of the approach to realistic noise levels and variations in experimental setting. Simulation of different rotor shaft geometries demonstrates general applicability of the approach with error levels similar to those in the PW615 engine. Guidelines on experimental setup, procedure and data processing are developed for future rotor forced-response experiment design and execution.

The thesis contributions are (1) a new approach for measuring rotor shaft dynamics in full aero-engines that enables the development of engine-condition based mechanical health monitoring and maintenance technologies, and (2) an extended reduced-order modeling framework that provides new capability for preliminary design of engine rotor systems and forced-response experimental design.