Impact of Aerodynamic Asymmetry on the Embedded Rotor Forcing and Mistuned Blade Response

Abstract This paper continues previous research by the current authors in a series of papers describing the impact of multi-row interaction on the forced response behavior of embedded compressor rotors. Specifically, this work dives into an aspect of the embedded rotor forcing that has been sparsely addressed in literature — the impact of an asymmetry in the upstream stator on the forcing function and blade vibration. Although the concept of introducing an asymmetry was introduced nearly 50 years ago, no papers in literature talk about the impact of multi-row interaction with an upstream asymmetric stator in place. All of the current authors’ previous work has discussed the impact of symmetric stators exciting a rotor upstream and downstream. This impact has been quantified at multiple torsional mode crossing some of which result in one-half of the asymmetric stator exciting the rotor and a crossing which results in the downstream stator exciting the rotor. Due to the inability of model reduction methods to model this phenomenon, full wheel computational domains were utilized for all simulations. A commercial code CFX was used for all simulations. In all of the multi-row cases investigated, only a single row excites the rotor. The additional downstream row serves as a reflecting wall that reflects physical waves, which leads to interference. The second section of this paper describes the details of the in-house mistuning code utilized to predict the blade responses of the individual rotor blades in the system. One of the inputs to this code is the modal force obtained using unsteady CFD in the paper’s first section. The in-house code is based on the FMM, which couples both aero and structural dynamics to predict the blade responses. The key conclusions from this study were: 1) An asymmetry in the upstream stator results in a significant reduction in the forcing function at the crossing frequency. However, the magnitude of forcing could remain significant with a shift to a ‘sideband’ frequency. 2) Introducing an asymmetry is beneficial only when the asymmetric stator excites the rotor and not when any of the other rows excites the rotor. 3) The impact of physical wave reflections was constructive at all torsional mode crossings in line with the earlier conclusions. 4) The mistuned response was well predicted by the FMM at all crossings despite the presence of significant blade disk interaction at these crossings.