Study On Characterization Of Debonding Defects In Thermal Barrier Coatings Based On Terahertz Technology

Thermal Barrier Coatings(TBCs)play an important role in the Thermal insulation protection of turbine engine blades,which can reduce the impact of extreme service environment caused by high temperature and high pressure on blade metal,and greatly improve the service life and energy utilization efficiency of turbine engine blades.The transverse crack propagation caused by thermal shock,the growth of thermal growth oxide and the fracture of the adhesive layer are easy to lead to TBCs debonding and reduce the service life of the thermal barrier coating.Effective monitoring and evaluation of the health status of TBCs to ensure the normal service of turbine engine has become an important problem to be solved.Terahertz(THz)technology is a new nondestructive testing technology,which can conduct nondestructive testing on TBCs structures in principle.In this thesis,a Terahertz(THz)technology based on the detection and analysis of TBCs debonding defects is proposed,which realizes the nondestructive noncontact testing of TBCs debonding defects.The main research work includes the following parts:(1)The Finite-Difference Time-Domain(FDTD)algorithm has been studied.The absorption boundary conditions,excitation source parameters,simulation step size and optical parameters of TBCs materials are determined.The FDTD model of terahertz signal with TBCs debonding defect is established.The simulation results show that the FDTD model is in good agreement with the analytical model,which verifies the correctness of the established model.(2)A terahertz detection and imaging method for thermal barrier coatings debonding defects was proposed.Firstly,the effect of the debonding defect on the terahertz signal is simulated by using FDTD model.It has been found that the debonding defect changes the terahertz reflection signal between the ceramic layer and the metal interface.And it can detect the debonding defect with a height of no less than5μm.Then the experimental study of the debonding defect was carried out.The terahertz signal is reconstructed and filtered by means of wavelet transform and wavelet filter.Then 2D and 3D images of defects are obtained by using the processed experimental signals.The experimental results show that the terahertz technique can be used to detect and image the debonding defects of thermal barrier coatings.(3)A classification method of thermal barrier coatings debonding defects based on machine learning was proposed.Firstly,the numerical model of FDTD is used to carry out the simulation research.The terahertz signals of different debonding defects were obtained,and the defect data set was constructed based on the signals.Then,the dimensionality reduction of the terahertz signal was processed by principal component analysis(PCA),and the characteristic signal of the debonding defect was extracted.Finally,the proposed classifier is used to classify the debonding defects.The results show that the performance of the SVM classifier is the best,the average accuracy is86.7%,and the running time is only 0.121 s.