航空航天蜂窝夹层结构服役环境恶劣,需要发展相应的结构健康监测技术以保证结构的安全运行.脱粘损伤因发生在内部胶层而难以监测,该文采用埋入胶层的分布式光纤传感器对蜂窝夹层结构脱粘损伤进行监测.在预埋损伤监测试验中采用高强光纤和边界浅槽的方式提高传感器存活效率,得到脱粘损伤应变特征.采用参数化建模方式对不同损伤位置进行大量有限元模拟,加入白噪声后得到含有脱粘损伤特征的13 500组数据.将模拟数据代入卷积神经网络进行训练,训练后的网络对试验损伤数据的识别准确率达到98.84%.该方法能够识别20 mm2 脱粘损伤,同时定位平均误差小于4 mm.
Debonding Damage Monitoring of Honeycomb Sandwich Structure Based on Distributed Optical Fiber
The aerospace honeycomb sandwich structure has a harsh service environment,and it is necessary to develop the corresponding structural health monitoring technology to ensure the safe operation of the structure.Debonding damage is difficult to monitor because it occurs in the internal adhesive layer.In this study,a distributed optical fiber sensor embedded in the adhesive layer is used to monitor the debonding damage of the honeycomb sand-wich structure.In the embedded damage-monitoring test,the use of high-strength optical fibers and shallow bounda-ry grooves improves the survival efficiency of the sensor and obtains the characteristics of the debonding damage strain.Using parametric modeling,a large number of finite element simulations were conducted at different damage locations.After adding white noise,13 500 sets of data containing debonding damage characteristics were acquired.Substituting the simulated data into a convolutional neural network for training,the trained network achieved an ac-curacy of 98.84% in identifying the experimental damage data.In the current study,this method can identify 20 mm2 debonding damage,with an average positioning error of less than 4 mm.
honeycomb sandwich panelstructural health monitoringdebonding damagedistributed optical fi-berconvolutional neural network
NETL
NSTL
万方数据
