A study on the stall detection of an axial compressor through pressure analysis

Abstract

In order to research the inherent working laws of compressors nearing stall state, a series of compressor experiments are conducted. With the help of fast Fourier transform, the amplitude-frequency characteristics of pressures at the compressor inlet, outlet and blade tip region outlet are analyzed. Meanwhile, devices imitating inlet distortion were applied in the compressor inlet distortion disturbance. The experimental results indicated that compressor blade tip region pressure showed a better performance than the compressor's inlet and outlet pressures in regards to describing compressor characteristics. What's more, compressor inlet distortion always disturbed the compressor pressure characteristics. Whether with inlet distortion or not, the pressure characteristics of pressure periodicity and amplitude frequency could always be maintained in compressor blade tip pressure. For the sake of compressor real-time stall detection application, a compressor stall detection algorithm is proposed to calculate the compressor pressure correlation coefficient. The algorithm also showed a good monotonicity in describing the relationship between the compressor surge margin and the pressure correlation coefficient.