Aeroelasticity of Turbomachinery

The aim of aeroelastic investigations on turbomachinery bladings is to provide procedures and tools for the prediction of flutter limits and forced response amplitudes due to aerodynamic excitation.

Experimental Investigations 

The development of modern aircraft engine compressors with increased pressure ratio and reduced weight has led to highly loaded stages in transonic flow. Turbomachinery blades are flexible structures. The relatively high flexibility of fan and low-pressure compressor blades can cause vibrations in frequencies which may be aeroelastically critical. Blade flutter is caused by interaction between the motion of the blades and the unsteady aerodynamic forces. These aerodynamic forces can either damp or excite the motion. In the latter case self-excited vibrations occur and the motion becomes unstable.

In order to avoid catastrophic engine damage due to flutter, design engineers have to examine the flutter susceptibility of the blading. Therefore, experimental and numerical tools are needed that are able to improve knowledge of the aeroelastic behaviour of the blading. In order to obtain precise insight into the aeroelastic behaviour of a vibrating blade assembly, a compressor cascade was designed for experiments near the flutter boundary. The cascade consists of 20 two-dimensional compressor blades. The blades were mounted onto elastic suspensions which allow for a torsional motion around mid-chord.

The experimental investigations were performed in the annular cascade tunnel at the EPFL (École Polytechnique Fédérale de Lausanne, Switzerland). The main feature of the wind tunnel is the possibility to investigate the aeroelastic behaviour of a cascade in a non-rotating test facility. A spiral flow is generated in order to simulate real inflow angles such as those that occur in rotating cascades.

In order to assess the aerodynamic stability of the blading, aerodynamic damping is determined by forcing the blades into controlled harmonic vibrations and by measuring the motion-induced unsteady pressure distributions. A Fourier transformation yields the first harmonic of the pressure at each transducer position and is related to the motion of this blade. The analysis of these data – in particular the out-of-phase unsteady harmonic pressure – gives insight into the local and global aerodynamic damping of the blades.

In the so-called "flutter experiments" the aerodynamic instability can overcome the structural damping, and the cascade starts to flutter  with self-excited vibrations. The increasing amplitudes and the phase shift between the blade motions are measured.

Current work regarding Forced Response in turbomachinery focuses on vibration modes excited by periodic, unsteady pressure fluctuations. These fluctuations always occur during operation since stationary blade rows interact aerodynamically with rotating blade rows, hence exciting vibrations of the elastic structure.  High cycle fatigue (HCF) failures can occur if vibration amplitudes increase above certain limits.

High-quality design tools and experimental validation data are necessary to avoid HCF failures and to give the blade designer more flexibility for increasing efficiency.. A European research project called Aeroelastic Design of Turbine Blades  (ADTurB) was initiated for the experimental investigations of forced response phenomena in high-pressure turbines (HPT). Partners in this project are industrial companies, universities, and national research organisations. The objectives of ADTurB are to understand these forced response phenomena and obtain high-quality validation data. To this end, a turbine stage was installed in the Wind Tunnel for Rotating Cascades (RGG) of the Institute of Propulsion Technology at DLR Göttingen. Three different rotors are equipped with unsteady pressure and blade vibration sensors.

Key components of the forced response experiments include measuring the aerodynamic forcing function acting upon the blades on the basis of blade row interactions and obtaining forced response amplitudes and vibration-induced aerodynamic damping with respect to realistic flow conditions, mechanical properties, eigenfrequencies, and eigenmodes.

Excitation at wake passing frequency of the periodic unsteady pressure fluctuations is a well-defined problem since the flow disturbance caused by each upstream (or downstream) blade row can be assumed to be identical.

If not, low engine order excitation, caused by small differences in the flow around the annulus, leads to vibrations of cascade modes with lower eigenfrequencies. For prediction of low engine order excitation response, experiments should yield knowledge of the vibration response with respect to the source of the annular variation and the sensitivity of the vibration response to these small changes.