Ultrasonic inverse identification of surface integrity of aviation functional coatings through material-oriented regularization

Abstract

Accurately controlling surface integrity of aviation functional coatings including defect, geometry, structure, and property is important to balance the comprehensive performance of aviation parts. In this paper, a new ultrasonic inverse method of surface integrity of aviation functional coatings is presented through a developed material-oriented regularization scheme, which combines model regularization, data regularization, and sensitivity matrix analysis. A sensitivity-matrix-based inversion method of ultrasonic detecting multiple surface integrity parameters is developed using the optimized high-sensitivity characteristic detection signal. Taking ultrasonic inverse detection of elastic constants of plasma-sprayed Al2O3 coating and inverse prediction of porosity of ZrO2 coating as examples, the validity of proposed ultrasonic inversion method is illustrated. Results show that the inverted elastic constants C11, C13, C33, and C44 of Al2O3 coating are 123.81 GPa, 34.38 GPa, 190.20 GPa, and 24.63 GPa, respectively. The relative error between inverted and experimental C33 and C44 are all less than 4.0%. The inverted elastic constants show obvious “inverse” elastic anisotropy, which is due to the preferred orientation of micro-cracks in Al2O3 coating. Compared with actual porosity of ZrO2 coating, the predicted porosity using sensitivity matrix optimized Gaussian process regression algorithm has a relative error of less than 5.3%. The proposed ultrasonic inverse method provides a new path for solving the complex ultrasonic non-destructive detecting problems, which are difficult for traditional “trial-and-error” ultrasonic methods.