Improved Galvanic Corrosion Model Considering the Corrosion Product Hindering Characteristic and Maximum Corrosion Depth

ABSTRACT Accurate corrosion prediction model can help to warn corrosion problems in advance. Therefore, to predict the galvanic corrosion process, this paper introduces the maximum corrosion depth, proposes the concept of corrosion product hindering characteristic k , and establishes a novel semi‐empirical galvanic corrosion model. The model shows different corrosion states under different k values and is suitable for corrosion processes under different environments and materials. Through the multiple sets of corrosion data obtained from the self‐piercing riveted joints of carbon fiber reinforced composites and 5052 aluminum alloy under various corrosion cycles, the model under experimental conditions is solved, and the prediction effects of the other three models are compared. The new model has a coefficient of determination R 2 = 0.9974, which is closest to 1, and MAE = 0.4973 is the smallest, so the model is considered effective.