Damage evolution and rupture time prediction in thermal barrier coatings subjected to cyclic heating and cooling: An acoustic emission method

Abstract

The real-time testing and quantitative assessment of damage evolution in thermal barrier coatings (TBCs) is desirable, but still intractable, especially at elevated temperature. In this paper, the fracture process of TBCs subjected to cyclic heating and cooling is monitored using an acoustic emission method. Based on the wavelet analysis of acoustic emission signals, damage modes in TBCs are discriminated. The results show that, due to thermal stress, it is preferential for vertical cracks in the heating stage and interface cracks in the cooling stage. The surface crack density and interface crack length are calculated to obtain the quantitative correlation of damage evolution in TBCs and acoustic emission parameters. The rupture time of TBCs can be predicted by the statistical analysis of acoustic emission signals.

Acoustic emission monitoring and damage mode discrimination of APS thermal barrier coatings under high temperature CMAS corrosion

“…In addition, four typical distribution histograms of energy coefficients for different damage modes are shown in Fig. 8 (t) (0.375-0.4375 MHz), obviously corresponding to the substrate deformation, surface vertical cracks, sliding interface cracks, and opening interface cracks, respectively [16,23,29]. These results demonstrate that the energy coefficient of wavelet packet is reliable to discriminate the damage modes of the TBCs attacked by the molten CMAS.…”

“…The frequency spectrum of AE signals are strongly dependent on the failure modes of materials, but are almost independent of the failure size and operating pressure [23,28]. Previous studies indicated that the surface vertical cracks, sliding interface cracks, and opening interface cracks correspond to the characteristic frequency bands of 0.20-0.25 MHz, 0.25-0.32 MHz, and 0.4-0.45 MHz, respectively [16,23,28,29]. After the statistical analysis on the characteristic peak frequency of AE signals, we also found three similar frequency bands for surface vertical cracks, sliding interface cracks, and opening interface cracks (Fig.…”

“…This method transforms the AE signals from time domain to frequency domain, and extracts the message reflecting the AE source [28]. The frequency spectrum of AE signals are strongly dependent on the failure modes of materials, but are almost independent of the failure size and operating pressure [23,28]. Previous studies indicated that the surface vertical cracks, sliding interface cracks, and opening interface cracks correspond to the characteristic frequency bands of 0.20-0.25 MHz, 0.25-0.32 MHz, and 0.4-0.45 MHz, respectively [16,23,28,29].…”

“…Besides, it is time-consuming and inconvenient to deal with each signal in this way. As a consequence, the wavelet analysis is applied to further analyze and identify the damage modes, because this analysis has good resolution in time and frequency domain for non-stationary random signals [23,29]. In wavelet analysis, a signal is divided into detail and approximate signal, the former is high-frequency components of the signal and the latter is the low-frequency part.…”

“…Trunova et al [22] reported the degradation evolution and failure mechanisms of TBCs during thermal cycling by analyzing microstructures and AE signals. Yang et al [23] investigated the thermal fracture behavior of TBCs under cyclic heating and cooling, and analyzed the correlation between damage evolution in the TBCs and AE characteristic signals. All these results provide theoretical and technical support for studying the failure mechanisms of the TBCs subjected to CMAS corrosion using AE technique.…”