Research On Detection Method Of Thermal Barrier Coating Thickness Based On Ultrasonic

With the rapid development of the aviation engines, inlet temperature of the frontof turbine engine increases, the turbine blades work in very poor environment withhigh-temperature and high-pressure for a long term. Thermal barrier coating is birth toprotect the blade from being oxidized by high temperature and corroded. On accountof the great differences between the physical properties and the thermal physicalparameters of the coating material and of the matrix material, coating may cause crack,hole, weak interface bonding, non-uniform thickness and other defects in themanufacturing process; the quality of the coating seriously affect the life andreliability of blade.This paper has established the theoretical spreading model according to theactual spreading of ultrasonic wave through the coating, aiming at the problems ofthat ultrasonic echo signal of the coating can’t directly identify the parametersassociated with the thickness of the coating in the time domain. Based on theinteraction relationship between the thermal barrier coatings and ultrasonic signals,ultrasonic signals are respectively analyzed from the time domain and frequencydomain, trying to dig out inherent information of characteristic parameters that caneffectively characterize the thickness of the coating, and using characteristicparameters to calculate the thickness of the coating. By comparing with the thicknessmeasured by scanning electron microscopy, the applicability and accuracy formeasuring thickness by ultrasonic nondestructive detection of these methods can beevaluated.By doing deconvolution and normalization processing on the signals of thecoating and the signals of the substrate in the frequency domain, the characteristicsignals of the coating are separated from the integrated response affected by manyfactors, the ultrasonic reflection coefficient amplitude spectrum is obtained by thecharacteristic signals of the coating is to characterize the thickness of the coating; hemutation time point of signals on the top and bottom surfaces of the coating aredetected by using wavelet transform modulus maxima and outstanding characteristicsof the singularity of the detection signals of Lipschitz index in the time domain, sothat the thickness of the coating can be calculated.The range of the thickness of the thermal barrier coating that this paper studies is 80~150μm, it is found by doing data fitting and correlation analysis on the datameasured by three methods: the results of measuring thickness by Lipschitz exponentfluctuations, however, the level reflected by the overall is the best, when obtainingthe average thickness of the coating, Lipschitz exponent should be used; the sum ofsquared residuals of WTMM analysis and spectral analysis s is relatively smaller,when obtaining the thickness of a single location, WTMM analysis and spectralanalysis should be used. Considering the sum of squared residuals, Pearsoncorrelation coefficient, adjusted goodness of fit and the slope, it can be known that thethickness measured by spectral analysis method is most reliable.