To measure thermal barrier coatings thickness at the high precision,a pulsed eddy current detection inversion method is proposed.A two-stage processing technique is proposed to preprocess the pulsed eddy current detection signals,and the principal component of the preprocessing signal is extracted by principal component analysis method.Finally,BP neural network is constructed to predict the coatings thickness.The COMSOL modeling and simulation experiment proves that the two-stage processing method can effectively reduce the lift-off effect and distinguish the variations of thickness characteristics.The feature extraction method based on principal component analysis can classify and identify the thickness variations of ceramic layer and bonding layer.The results show that the average relative error of this method is about 0.4％for the thickness of ceramic layer and about 2.6％for the thickness of bonding layer.It can be seen that the inversion accuracy of thermal barrier coatings thickness by the mentioned method is higher.
国家科技期刊平台
NSTL
万方数据
