Terahertz Diagnosis Mechanism And Experimental Investigation Of Thermal Barrier Coatings Failure

Thermal Barrier Coatings(TBCs)on the aero-engine turbine blades is an important factor to improve the operating efficiency of the engine.The thickening of thermally grown oxide(TGO)caused by oxidation at high temperature and the propagation of microcracks are the main factors causing the failure of TGO.At present,the most common detection method for thermal barrier coatings failure is to use scanning electron microscope(SEM)to obtain the section of thermal barrier coatings under static conditions.However,it is a destructive testing method,which is different from the actual working conditions of thermal barrier coatings.Terahertz Pulsed Imaging(TPI)has become a new potential method of TBCs nondestructive detection and quantitative evaluation with high signal-to-noise ratio,security and high detection depth.The THz diagnosis of the failure processes of TBCs was investigated in this dissertation.In particular,the terahertz wave propagation model was established;highprecision signal analysis methods were explored in the project;the surface crack,internal debonding and oxide layer growth process in the turbine blades were monitored to achieve the non-destructive evaluation and prediction of failure processes in TBCs.The main contents are as follows:The theory governing the interaction between THz waves and substance was analyzed to further study the propagation mechanism of terahertz waves in the thermal barrier coatings.A finite element model on THz wave propagation in the TBCs was established;a parametric study was systematically carried out to evaluate their influences on TBCs,such as excitation source and boundary conditions of the model.Subsequently,the model was employed to conduct forward analysis on terahertz wave propagation.The signal characteristics of each layer boundary of thermal barrier coatings were obtained.The multi-layer coating analysis of thermal barrier coatings was realized by using frequency domain deconvolution and the stationary wavelet transform(SWT).The results obtained by using these two signal processing methods were compared to extract the coating thickness.The conclusion can be drawn that the SWT is a better signal analysis method than frequency domain deconvolution method for TBCs.As the TPI system presents a limited axial resolution for its limited wavelengths,the influence of the bandwidth and the device,especially when measuring a very thin thermally grown oxide in TBCs whose thickness is close to 10 μm(coating will start peeling).An SWT-BP algorithm combining both the SWT and the backpropagation(BP)neural network was proposed.The algorithm made full use of the advantages of the SWT algorithm to extract detailed information and the powerful predictive ability of the multi-layer neural network BP algorithm,the prediction results were in good agreement with the real results and the proposed algorithm was able to achieve a thickness prediction below 10μm of TGO and to improve the resolution of terahertz detection.The proposed algorithm is suitable for thin thickness detection of the TGO.Two tomography methods including TPI and Optical Coherence Tomography(OCT)were combined to evaluate the typical defects of TBCs(i.e.;interfacial cracks,internal debonding,surface high-temperature cracks,etched cracks,and spallation cracks etc.).TPI system has higher penetration depth than OCT,hence it can be used to nondestructively detect and evaluate the interfacial cracks,as well as the internal debonding defects in the sample.On the other hand,axial resolution of the near-infrared(NIR)OCT system is at micron or submicron level,can detect more detailed information;while owning to the strong absorption and strong scattering of TBCs in NIR range,its detection depth is limited.Therefore,OCT is superior in detecting surface cracks of TBCs.Based on the complementary advantages of TPI and OCT,a method of nondestructive testing and quantitative evaluation of TBCs using TPI and OCT was proposed in this dissertation.In order to deeply study the growth law of oxide layer thickness under cyclic thermal load,a finite element model of TGO was established.The simulation results were basically consistent with the actual experimental coating thickness,which verified the validity of the model and provided theoretical support for the early thickness measurement of TGO.From 0-140 h,results demonstrated that the oxidation growth with increase in high temperature oxidation time,whilst the oxidation rate slowed down.Since the thickness of TGO remained below 10 microns within a certain period of time,the proposed method of solving the integral refractive index of TBCs and the SWT-BP method were used to analyze the samples with the TGO of different thicknesses(under the ceramic layer).In the dissertation,the diagnostic mechanism and experiments of TBCs failure were investigated.The resolution of the terahertz detection system was improved.The monitoring problem of the growth process of early TGO in the thermal barrier coating was solved;also,it provided theoretical and experimental support for the nondestructive detection and quantitative evaluation of the early failure of thermal barrier coatings on turbine blades.