Experiments with vibrating cascades are done in the annular channel (RGP).
The annular channel has several advantages for vibrating cascade experiments:
Most vibrating cascade experiments in the annular channel are done in a forced vibration mode. This means that the blades are forced to oscillate in a desired vibration mode. For controlled forced excitation experiments it exists an excitation and control system at the LTT. The system is configured for 20 blades. The blades are forced into vibration by means of electromagnetic exciters. The blade vibration is measured by inductive displacement transducers (model TQ 102) from the company VIBRAX (Switzerland). These transducers are fixed on an impact ring. The impact ring (mechanical brake) is controlled axially by a hydraulic system. It permits to limit the blade vibrations when the blades oscillate in the self-started flutter mode, in order to prevent the springs from breaking. An electronic control system allows to establish and maintain an organized vibration mode of the cascade. Each blade has its own feedback control loop. The blade vibration amplitudes can be selected individually and almost any cascade vibration mode can be realized. However such an electromagnetic system can exert only a limited force on the blades. Therefore, the blade vibration amplitudes and phase angles are not rigorously constant but subject to fluctuations that depend on the mechanical properties of the cascade (such as coupling of the blades) and the vibration mode selected as well as non periodic flow perturbations. The excitation frequencies have to be close to the eigenfrequencies of the blade-mass-spring system.
 |
A cascade for vibrating blade measurements consists of the following pieces: an annular core, the spring elements, the mass elements, the blade bases and the blades. The core holds the spring elements in conical borings. The shape and stiffness of the spring elements determines the eigenfrequencies of the cascade. The cascade can only be excited close to the eigenfrequencies of the blades as the excitation system yields only a limited force. For that reason the spring elements must be manufactured with a high accuracy. Transformer plates are fixed to the mass elements. The electromagnetic forces act on these plates. By using different materials (plastics, aluminum, steel) and shapes for the mass elements the eigenfrequencies can be varied in a certain range but to a lesser extent as with the springs. The blades are fixed to the blade bases. They have overlap in order to reduce the flow leakage into the core of the annular facility but they must also permit a free movement of the blades at all flow conditions while the cascade is vibrating. Assembled the blade bases form a ring. In order to reduce the mass they are made of aluminum. The blades are manufactured by means of electroerosion. |
A vibrating blade segment
The assembled segments form a vibrating cascade.
Usually the cascade is instrumented with classical pressure tabs for steady-state measurements at different loacations on the blades and in the channell wall. For the unsteady pressure measurements three types of pressure transducers have been used until now: transducer mounted on ceramic pellets (ENDEVCO 8515B-50), cylindrical pressure transducers (ENDEVCO 8514-50) and subminiatur pressure transducers (KULITE LQ 1-062-25A).
The ENDEVCO transducers are mounted inside the blades and connected to the blade surface by pressure tabs. This results in a smooth surface but the dead volume and the pressure tab may alter the response of the transducers to pressure fluctuations in the flow. These effects are usually negligible but they increase with frequency. The flat KULITE transducers are surface mounted. They can be placed closer together (< 3.5mm) and closer to the trailing edge than the other transducers. Their disadvantage is that they are exposed to all possible damaging effects (particles in the flow, mistreatment, etc.) and that they disturb the smoothness of the blade surface. This may be critical in the trailing and leading edge regions. The signal of each transducer is amplified, filtered and digitized. The raw data are stored on a data acquisition PC and evaluated. The present data acquisition system permits a sampling rate of 180000 samples per second and the recording of up to 64 channels. The sampling rate per channel depends on the connected number of channels.
In order to examine and to compare the unsteady pressures one has to get information about the steady-state flow properties. The pressure distribution is measured on the blade surfaces. Thus, one obtains for example the isentropic Mach number along the chord of a blade.
The inlet and outlet flow properties are measured with steady-state flow-probes.
This yields 2-D information about flow-angles, Mach-numbers, total pressures
and the mass-flow distribution in front of and behind the cascade.
Paint flow visualization is used to get an idea about the flow patterns at
the channel walls and on the airfoil surfaces.
(AVI-video: 4.166 kb !)
(AVI-video:
5.034 kb !)
Unsteady aerodynamic pressures are measured with the cascades vibrating in single blade vibration mode (only one blade vibrates at a time) or in travelling wave mode (all blade vibrate with a constant interblade phase angle). The time-dependent pressures are measured with piezoelectric pressure transducers that are mounted inside the blades and at the outer channel wall. An ensample averaging technique is used to calculated the time-dependent pressures at a specific transducer location. An integration of the time-dependent pressures over the blade surfaces yields the unsteady forces and thus tells if the unsteady pressures tend to further increase a blade vibration (self-excitation) or if the cascade vibration is damped (->damping coefficient). Unsteady aerodynamic influence coefficients for each vibrating blade can be calculated from the measured data of different cascade vibration modes using a decomposition technique.
Examples of the visualization of test results:
Comparison
of unsteady pressure levels and isentropic Mach number isolines measured at
the outer channel wall.
A
cube of colour coded unsteady pressure coefficients on a blade surface with
the chordwise position, the interblade phase angles and the angle of attack
as the three axis parameters for a constant inlet Mach number.
Animation
of measured time-dependent pressures at the outer channel wall while the cascade
is vibrating in travelling wave mode (s=180°).
(AVI-animation: 430 kb !)
Animation of measured time-dependent pressures at the outer channel wall.
The middle blade is vibrating in single blade vibration mode.
(AVI-animation: 347 kb !)
In the frame of a european research project (Brite-Euram ...) a rotor with elliptical struts will be put in front of a turbine cascade in the annular channel in order to examine the influence of wakes on vibrating cascades. The wakes create periodic time-dependent disturbances. The unsteady forces caused by the wakes excite the blades to vibrate and the blade vibrations cause additional unsteady aerodynamic forces. One coal of the project is to study the superposition and interaction of the unsteady aerodynamic forces caused by the wakes and those created by blade vibration.
Calculated Mach number distribution in a turbine cascade with rotating elliptical
struts in front.
(AVI-animation: 376 kb !)
