News Center

Home

Current Status and Development Trends of Research on Condition Monitoring of Aeroengine Spindle Bearings (II)

Release time:

2023-01-13

Source:

Journal of Aerodynamics

Due to its high load and wide range of operating conditions, the main shaft bearing has a short time from showing damage characteristics to failure. Therefore, once fault symptoms are found, the operating conditions should be decisively adjusted and maintenance should be arranged as soon as possible. According to the different types of monitoring signals, sensors are divided into vibration sensors, acoustic sensors, acoustic emission sensors, etc.

2. Main status monitoring methods for spindle bearings

2.1 Vibration feature monitoring methods

When spindle bearings experience fatigue, wear, and other faults, abnormal vibrations may occur. The vibration monitoring method is achieved by installing vibration sensors in appropriate positions on the bearing seat or box to collect signals and analyze them.

The installation position of vibration sensors is limited by the engine structure, and usually only one vibration sensor is installed at the gearbox of the aircraft engine. Moreover, the aircraft engine system has problems such as long vibration transmission path, complex frequency components, and severe signal attenuation, which puts forward high requirements for the analysis method of vibration signals [34].

Chen Guo et al. [35] studied the sensitivity issue of spindle bearing fault diagnosis based on the measurement point signals of the gearbox. When the connection stiffness between the rolling bearing and the gearbox is small, the vibration signal will have a significant attenuation. However, by selecting appropriate methods, more accurate diagnosis can still be made; Zhang Xiangyang et al. [36] proposed a bearing fault diagnosis method based on Convolutional neural network for the vibration signal of the gearbox, and the effectiveness of this method was proved through experiments.

Diagnosis can be carried out by calculating time-domain characteristics, frequency-domain characteristics, and time-frequency domain characteristic parameters of vibration signals [37 ⁃ 38]. Time domain parameters include dimensioned parameters such as effective value, root mean square value, peak value and dimensionless parameters such as kurtosis, Crest factor, waveform factor and margin index; The frequency domain analysis method includes power spectrum, amplitude spectrum, cepstrum, complex cepstrum, high-order spectrum, envelope spectrum, etc; Time frequency methods include short-time Fourier transform, Wigner Ville distribution [39], spectral kurtosis [37,40], wavelet analysis [41], Stochastic resonance [34,42], etc.

Scholars at home and abroad have proposed many fault diagnosis methods based on the vibration signals of aviation engine spindle bearings, which have been proven to have high practicality and accuracy through experiments and other methods. Zhang et al. [43] proposed an AMS (alternative minimization solver) ⁃ CluLR method, which was demonstrated through high-speed tests of aircraft engine bearings to accurately identify outer ring faults of bearings; Wang et al. [44] proposed a quantitative diagnosis method for early faults of high-speed bearings in aircraft engines based on support vector machines, which can distinguish different types of faults and different degrees of the same type of fault; Liao Mingfu et al. [45] found that the frequency spectrum of the vibration signal of the intermediate bearing in an aircraft engine produces a frequency doubling "constant spacing" feature that does not change with speed. Through experiments, it has been proven that this feature can serve as a basis for fault diagnosis.

Due to the convenience of collecting vibration signals, relatively low sensor prices, and mature theories, various rolling bearing monitoring systems both domestically and internationally are mostly developed based on vibration sensors.

2.2 Acoustic feature monitoring methods

Essentially, sound is composed of vibrations