摘要
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 rotor blade of aero engine, the first resonant frequency is gotten using swept method at the condition of whorl pool vibration. The first resonant frequency of different damage location and damage radius is obtained by prefabricating different damages on the air flow side of the blades. The non - linear relationship between damage and evaluation parameters is constructed based on the BP neural network and then the quantitative assessment is presented primarily, and the forecast of damage has been processed.
出处
《汽轮机技术》
北大核心
2011年第4期277-279,共3页
Turbine Technology
基金
总参维修预研基金(5132701)资助
关键词
损伤度
BP神经网络
叶片
固有频率
damage degree
BP neural network
blade
resonant frequency