Intelligent fault diagnosis of rotors always requires a large amount of labeled samples, but insufficient vibration signals can be obtained in operational rotor systems for detecting the fault modes. To solve this problem, a domain-adaptive transfer learning model based on a small number of samples is proposed. Time-domain vibration signals are collected by overlapping sampling and converted into time-frequency diagrams by using short-time Fourier transform (STFT) and characteristics in the time domain and frequency domain of vibration signals are reserved. The features of source domain and target domain are projected into the same feature space through a domain-adversarial neural network (DANN). This method is verified by a simulated gas generator rotor and experimental rig of rotor. Both the transfer in the identical machine (TIM) and transfer across different machines (TDM) are realized. The results show that this method has high diagnosis accuracy and good robustness for different types of faults. By training a large number of simulation samples and a small number of experimental samples in TDM, high fault diagnosis accuracy is achieved, avoiding collecting a large amount of experimental data as the source domain to train the fault diagnosis model. Then, the problem of insufficient rotor fault samples can be solved.

The vibration characteristics of bolted structures are crucially affected by the pre-tightening torque. An approach for identifying the pre-tightening torque dependent stiffness parameters of bolted joints is proposed in this paper. Firstly, the interface of the bolted joint is characterized by the thin layer element with the isotropic material property, and the parameter value of the property is assigned relative to the distance from the center of the bolt; the influence of the bolt is ignored. Secondly, the model updating method is adopted to identify the parameters of thin layer elements using experimental data, and modal data under different values of pre-tightening torque in the range of 2 N·m~22 N·m are obtained; the torque wrench is used to determine the pre-tightening torque in the modal test. Finally, after identifying the material parameters using partial experimental data on pre-tightening torque range, the empirical equation of the interface parameters with the pre-tightening torque parameter is obtained by curve fitting and the rest of the experimental data are used to verify the accuracy of the fitted empirical equations. It is concluded that this method can obtain all the parameters of the equivalent thin layer elements within a certain range of pre-tightening torque, which can provide a reference for the empirical modeling of bolted structures, improve modeling efficiency and reflect the characteristic performance of real structural dynamics.

A multi-scale fatigue analysis method for braided ceramic matrix composites (CMCs) based on sub-models is developed in this paper. The finite element shape function is used as the interpolation function for transferring the displacement information between the macro-scale and meso-scale models. The fatigue failure criterion based on the shear lag theory is used to implement the coupling calculation of the meso-scale and micro-scale. Combining the meso-scale cell model and the fatigue failure criterion based on the shear lag theory, the fatigue life of 2D SiC/SiC is analyzed. The analysis results are in good agreement with the experimental results, which proves the accuracy of the meso-scale cell model and the fatigue life calculation method. A multi-scale sub-model fatigue analysis method is used to study the fatigue damage of 2D SiC/SiC stiffened plates under random tension–tension loads. The influence of the sub-models at different positions in the macro-model element on the analysis results was analyzed. The results shows that the fatigue analysis method proposed in this paper takes into account the damage condition of the meso-structured of composite material, and at the same time has high calculation efficiency, and has low requirements for modeling of the macro finite element model, which can be better applied to the fatigue analysis of CMCs structure.