Fabrice Rottmeier
An experimental investigation has been conducted in the non-rotating annular test facility of the "Laboratoire de Thermique Appliquée et de Turbomachines" (LTT), "École Polytechnique Fédérale de Lausanne" (EPFL). During this investigation, the unsteady aerodynamic response of a turbine cascade was investigated for three different cases: (1) the clamped blades subjected to periodic, upstream generated aerodynamic gusts, (2) the cascade forced to vibrate in the travelling wave mode in a uniform flow, and (3) the cascade forced to vibrate in the presence of the upstream generated aerodynamic gusts, with a common excitation frequency and a constant gust-vibration phasing. These measurements were aimed at identifying important aspects of the unsteady aeroelastic behavior of the blades. Particular attention was focused on the relationship between the time-dependent flows of (1) and (2) and that of (3) in an effort to better understand aerodynamic forced response phenomena in turbomachinery The experimental tasks detailed above have been performed in a non-rotating annular test facility using a test rig composed of rotating wake generators and a fixed turbine cascade. There are two rotating wake generators: one possesses 13 elliptical struts and the other has 22. Each one is designed to generate wake profiles similar to a blade row. The turbine cascade consist of 20 blades attached to separate torsion suspension system, with a magnetic excitation feature, allowing the control of the cascade's vibration mode. Several blades are instrumented to measure the resulting surface steady pressures, unsteady pressures and blade vibration mode The specific objectives of this investigation were: measure the unsteady aerodynamic blade response to: (1) imposed cascade vibration modes (travelling wave mode), (2) upstream generated aerodynamic gusts, and (3) combinations of these two effects. using the above experimental results, address the local validity of the assumptions inherent in linearized treatments of the forced response problem. Specifically, can the local unsteady blade loading be considered as a linear superposition of the unsteady forces derived individually from the cascade's vibration mode and from the aerodynamic gusts? The final conclusions of this work were: the simultaneous measurements have demonstrated the important influence of the gust-vibration phase angle Φ on the blade surface time-dependent pressure distribution and has identified it as the key parameter. For a given test configuration and flow condition, the selection of the gust-vibration phase angle allows a local constructive or destructive interaction between the main harmonics of both excitation sources. Generally, this suggest that the gust-vibration phase angle has an important effect on the excitation levels and excited modes of the cascade. The practical implication of this phenomenon is that the judicious choice of a gust-vibration phase angle can diminish significantly the aerodynamic excitation levels for a given cascade vibration mode, flow condition and test configuration of a single stage. It is even possible that certain periodic variations of this phase angle produces another excitation source in turbomachinery. the simultaneous vibrating cascade and upstream generated aerodynamic gusts excitations can be accurately predicted by the linear superposition of the individual gust-induced and vibration-induced unsteady flow fields. This principle was shown to be applicable locally for various test configurations, including different engine orders, inter-blade phase angles, gust-vibration phase angles, and flow conditions. The only significant discrepancies were observed in the presence of shocks, but these were limited to a localized region reducing their importance in terms of the overall unsteady aerodynamic loading of the blade. It has been shown that these discrepancies were mainly due to the variation of the shock's mean location between gust-response only, controlled-vibration only and simultaneous measurements. The main practical advantage of this linear superposition principle is that the numerical and experimental investigation of the aerodynamic forced response problem can be separated into: (1) the identification of the local forcing-function due to upstream generated gusts only, (2) a local aerodynamic stability analysis of the cascade alone. The simultaneous measurements have shown that measurement errors vary significantly as the local pressure disturbance contributions due to the gusts and the cascade's vibration mode interact constructively or destructively. Particular attention is needed in simultaneous measurements in order to separate measurement error effects and physical phenomena.
EPFL2003
