Crack propagation monitoring of rotor blades using synchroextracting transform

Blade fatigue fracture is a serious aero-engine fault that may lead to an aircraft crash. It is essential for ensuring the safety of the aero-engine operation to monitor the blade crack propagation. Generally, the crack propagation of rotor blades can be detected by the natural frequency shift. The time-frequency analysis method with high time-frequency resolution can detect the natural frequency shift during crack propagation. In this paper, the synchroextracting transform with the advantage of high time-frequency resolution is proposed to extract the time-frequency feature of the vibration signal of the cracked blade. In simulation, the crack length is defined to increase exponentially. The blade finite element model is used to generate the simulated signals. Compared with short-time Fourier transform, synchroextracting transform with the advantage of high time-frequency resolution can detect the natural frequency shift of the cracked blade earlier. Furthermore, Chebyshev window is used to solve the problem of spectral leakage caused by an interval sampling strategy. In a spin testing, two blades without artificial cracks suddenly broke. Strain signals are used to verify the effect of time-frequency resolution on crack propagation monitoring. Short-time Fourier transform is difficult to monitor the crack propagation in the spin testing because of the interval sampling strategy. However, synchroextracting transform with the advantage of high time-frequency resolution can detect the crack through the natural frequency shift of the cracked blade. Finally, in order to reveal the damage of the cracked blade, the crack length is estimated qualitatively by combining the natural frequency and the rotational speed measured in experiment with the results of finite element analysis. The feasibility of the crack estimation method is verified by the experiment in a non-rotating case. Blade fatigue fracture is a serious aero-engine fault that may lead to an aircraft crash. It is essential for ensuring the safety of the aero-engine operation to monitor the blade crack propagation. Generally, the crack propagation of rotor blades can be detected by the natural frequency shift. The time-frequency analysis method with high time-frequency resolution can detect the natural frequency shift during crack propagation. In this paper, the synchroextracting transform with the advantage of high time-frequency resolution is proposed to extract the time-frequency feature of the vibration signal of the cracked blade. In simulation, the crack length is defined to increase exponentially. The blade finite element model is used to generate the simulated signals. Compared with short-time Fourier transform, synchroextracting transform with the advantage of high time-frequency resolution can detect the natural frequency shift of the cracked blade earlier. Furthermore, Chebyshev window is used to solve the problem of spectral leakage caused by an interval sampling strategy. In a spin testing, two blades without artificial cracks suddenly broke. Strain signals are used to verify the effect of time-frequency resolution on crack propagation monitoring. Short-time Fourier transform is difficult to monitor the crack propagation in the spin testing because of the interval sampling strategy. However, synchroextracting transform with the advantage of high time-frequency resolution can detect the crack through the natural frequency shift of the cracked blade. Finally, in order to reveal the damage of the cracked blade, the crack length is estimated qualitatively by combining the natural frequency and the rotational speed measured in experiment with the results of finite element analysis. The feasibility of the crack estimation method is verified by the experiment in a non-rotating case.

Automated summary

Blade fatigue fracture is a serious aero-engine fault that requires monitoring blade crack propagation to ensure safety. Utilizing synchroextracting transform with high time-frequency resolution can detect the natural frequency shift of cracked blades earlier, which is crucial for crack propagation monitoring.