Abstract: Corrosion morphology image is one of the most important features for the evaluation of the corrosivity of a material. By means of an optical microscope, the corrosion morphology images as well as the pit depths of LY12CZ samples exposed to experimental environments are captured. After digital image processing, the color types of these images are transformed, and then three kinds of corrosion morphology characteristics are extracted. By using a GM (1,3) grey model, the relationship between the features of corrosion morphology images and pit depths are studied, and a model based on the morphology characteristics is established which can predict the pit depth. Based on this model, a fracture mechanics method which treated pitting as cracks is established by the software AFGROW, which is used to predict the fatigue life of LY12CZ samples. The results show that there is good agreement between the measured pit sizes and the predicted data based on the grey model. Moreover, the fatigue lives calculated based on corrosion morphology characteristics are reasonable as compared with the results calculated according to experimental data, the average relative error being 8.84%.

Key words: corrosion image, digital image processing, corrosion morphology characteristic, corrosion damage, fatigue life, grey model, aluminum alloys