针对发动机叶片损伤的评定,通过低压压气机二级转子叶片的涡流激振试验,在叶片振动状态下用扫频法测试叶片一阶固有频率;在叶片进气边不同位置预制不同半径的损伤,得到了不同损伤位置、损伤半径下叶片一阶固有频率,并利用BP神经网络方法对损伤和评定参数之间的非线形关系进行了评定,初步验证了理论分析结果,进行了损伤预测。 In order to evaluate the damage of the engine blade, the first-order natural frequency of the blade was measured by swept-frequency method under the vortex vibration of the second-stage rotor blade of the low-pressure compressor. Different radial damage was prefabricated at different positions of the blade inlet, The first order natural frequency of the blade under different damage locations and damage radius was obtained. The nonlinear relationship between damage and evaluation parameters was evaluated by BP neural network method. The theoretical analysis results were preliminarily verified and the damage prediction was carried out.