Image recognition for lifetime prediction of organic coatings in the deep-sea environment

摘要

The rapid failure of organic coatings in deep -sea environments complicates accurate lifetime prediction [1, 2]. Given the rapid cracking characteristic on the coating surface in this environment, a comprehensive “performance -structure” failure model was established. Image recognition technology is used to extract information from SEM images of epoxy mica coatings serving for different periods of time in the simulated deep-sea fluid -pressure environment. A targeted approach containing convolutional neural networks and post-processing has been established for the crack area detection of coating surface. The length distribution and the statistical evolution of cracks were summarized, to obtain the kinetic equation of the cracks related to coating structure degradation. Based on this achievement, a comprehensive failure model combining coating properties and coating structure degradation is developed. The relative weights of three dominant factors in coating failure, including water diffusion, coating adhesion, and crack length were calculated by the gray relational analysis method. A relatively accurate prediction of coating lifetime was performed through the established “performance-structure” failure model.