Unsteady Aerodynamic Measurements on a Rotating Compressor Blade Row at Low Mach Number

Abstract

An experiment was conducted on a heavily instrumented isolated model compressor rotor to study the unsteady aerodynamic response of the blade row to a controlled pitching oscillation of all blades in an undistorted flow, and to a circumferential inlet flow distortion with nonoscillating blades. To accomplish this, miniature pressure transducers were embedded in the blades and the unsteady pressure time histories were recorded. Both phases of the experiment were performed over a wide range of flow coefficient, from C x /U m = 0.6 to 0.95 in 0.05 steps, and data were taken at each condition for sinusoidal disturbances characterized by one, two, and four per revolution waves. Steady-state data were acquired for flow coefficients from 0.55 to 0.99 in 0.05 steps. In this paper the steady and unsteady results of the portion of this experiment dealing with oscillating blades are compared with analytical predictions, and the steady results are compared with experimental data from previous work. Although the model blades were instrumented at five spanwise stations, only the midspan measurements will be presented herein. The measured pressures for nonoscillating blades were in good agreement with the steady potential flow predictions (and with previous steady experimental data) when the measured exit angle was imposed as the downstream boundary condition for the analysis. It was found that a quasi-steady approach yielded marginally acceptable agreement with the experimental results for the lowest frequency tested. For the higher reduced frequencies, the experimental data could not be modeled in this manner. In contrast, a comparison of the measurements with the Verdon–Caspar unsteady potential flow theory produced generally good agreement except near the leading edge at high mean incidence (i.e., at low flow coefficient). At high incidence the blades in this experiment had very high steady pressure gradients near the leading edge and it is suspected that this may be responsible for the lack of agreement. The agreement was somewhat better at the higher frequencies.