Inverse Analysis Of Acoustic Emission Source Of The Damage In Thermal Barrier Coatings

Thermal barrier coating (thermal barrier coatings, TBCs) are widely used in aviation turbineengines because of its excellent thermal insulation, wear resistance and corrosion resistance.However, because of its complex structure, poor service environment, there will be cracking anddelamination at the interface between the surface of the coating and thus eventually lead to coatingspalling failure, even catastrophic accidents. Because there are several factors which can give riseto the surface cracking and delamination of the interfaces, It is very difficult to predict the timewhen the surface and interface cracks initiate, the location of crack and the degree of damageowing to the cracks initiation and propagation along the way perpendicular to the surface coatingand parallel to interfaces in TBCs. The surface and interface cracks are two primary modes whichlead to the failure of TBCs, it has been demonstrated experimentally that the main frequency ofsignal of acoustic emission can separate the two failure models. However, the method todistinguish the crack models in just verified in experiment, there is lacking theoretically explainsfor supporting the method scientifically. In this paper, we monitored real-time the failure processof thermal barrier coatings under bending test by AE and acquired the signals from the surface andinterface cracks, calculated the Green function by the generalized ray method, and demonstratethat it is reasonability and scientifically theoretically by using acoustic emission to distinguishsurface crack and interface crack, the major research content is as follows:First, we established the mathematical relationship between the moment tensor ofmathematical model of the crack in thermal barrier coating and the acoustic emission signal. Thesurface and interface crack will produce elastic stress wave, after many times of reflection,refraction and transmission before reaching the acoustic emission sensor and the elastic wavereceived by acoustic emission sensor are converted to electrical signals output. According to therelationship among moment tensor, the Green displacement function and acoustic emission signals,we finally established the direct relationship between the moment tensor related with crackinformation and acoustic emission waveforms.Second, we calculated the transfer function of the acoustic emission sensor by pencil leadfracture experiment. The transfer function represents the characteristics of the acoustic emissionsystem, nothing to do with the input. Acquiring the acoustic emission signals generated excitatedby point force from the pencil lead breaking experiment and using the generalized ray method tocalculate the Green displacement function where the sensor were placed on, the transfer functionof sensor is finally obtained by inverse convolution. Third, we monitord the failure process of thermal barrier coatings in which the surface andinterface cracks initiate and expand under bending test with acoustic emission, gathered thesignals from cracks and used the generalized ray method to calculate the corresponding Greenfunction in the half space. With the collected acoustic emission signal, the time function aboutsurface and interface crack and the frequency distribution of surface and interface crack, thefrequency distribution about surface crack is about0.4-0.6MHz, the frequency distribution aboutinterface crack is about0.8-1.1MHz. Thus It has been proved that it is scientific and reasonable todistinguish thermal barrier coatings failure behaviors by acoustic emission frequency.