Forced Response Due to Vane Stagger Angle Variation in an Axial Compressor

Abstract

Recent developments in turbomachinery design require an improved prediction accuracy of blade vibrations to maintain safe operations. This article aims to investigate the accuracy of numerical aeroelastic approaches for the calculation of blade vibrations. For validation, extensive aerodynamic and forced response measurements in an 1.5-stage axial compressor with a blade integrated disk (Blisk) are presented. The excitation intensity of the vibration is controlled by varying the stagger angle of the inlet guide vane (IGV). In addition, a second engine order is imposed by a nonsymmetric circumferential vane angle distribution to simulate a multistage behavior. Experimental validated Reynolds-averaged Navier–Stokes (RANS) simulations in both the frequency and the time domain are compared to assess the prediction accuracy of the numerical approaches. The numerical results agree with the experiments for low and intermediate vane angles. However, at high IGV stagger angles and when exciting multiple engine orders, the inaccuracy in the prediction of flow separation by the RANS simulations leads to an overprediction of vibration amplitudes. This exaggeration becomes even more pronounced in the frequency domain simulations. Time domain methods with a time lag formulation tend to be efficient and more accurate approaches for large separated flow regimes.