Research On The Detecting Technology Of Unstable Signal In The Compressor

The instability of aeroengine mainly refers to its key component, the compressor’s instability. So far, the study on the mechanism shows that the instabilities are caused by the development of stall cells which are generated from tip or middle of the blades. They mainly display in two forms, rotating stall and surge. Rotating stall and surge may occur simultaneously or individually. They all belongs to the pneumatic unstable phenomenon inside the compressor. They restrict the promotion of the efficiency for engine, especially cause violent vibration of the compressor blades, resulting in the damage of mechanical structure or even the air parking.The detecting of unstable signal in the compressor means first analyzing and processing the unstable inception mixed in the pressure signal, then extracting the instability characteristics, and taking proper prediction algorithm and instability criterion to give instability warning in the end. It is the foundation of preventing instability of the compressor, and it is also an important part of the active control system.Compressor model is used to describe the dynamics of pressure ratio, the average mass flow and the disturbances of mass flow. It is the basis for analyzing and judging stall and surge, and also for designing active control system. Based on the MG model, we established the mathematical model of unstable signal in the compressor under the condition of non-constant speed, giving the detailed derivation and the simulation results. The results show that the model can well describe the instability phenomenon (rotating stall and surge) in the axial compressor. It can be used for stability analysis of the axial compressor system, prediction of stall and surge and design of the active control system. The key to detecting technology of unstable signal in the compressor is the detecting method. In this paper we introduce the method based on the pressure signal at the outlet and the method based on pressure signal at the first stage, and also discuss the wavelet variance algorithm and the correlation integral algorithm. Then we analyze the two algorithms based on the dynamic model and the trial test data of compressor. The results show that both of the two algorithms can effectively detect the unstable signal and give instability warning. However, they are different in universality, accuracy and timeliness. They also have complementary advantages and disadvantages.Finally, according to the technical requirements of the project, we propose a scheme for the detecting system of unstable signal in the compressor. The main content includes:the overall scheme, the detecting algorithm, the selection of pressure sensors and the system architecture. The scheme will serve as the basis of the late-stage research, guiding the hardware and software realization of the detecting system of unstable signal in the compressor.