A computational framework for probabilistic modeling of galvanic corrosion for automotive applications

Abstract To address the need for reduced vehicle weight and improved environmental sustainability, the automotive industry has increasingly turned to mixing lightweight materials and alloys with metal alloys. However, this integration of dissimilar materials has heightened the risk of galvanic corrosion. This study addresses the gap in modeling of galvanic corrosion under dynamic thin film electrolyte by incorporating data derived from real-world weather conditions and finite element simulations. The presented model successfully captures the trend of galvanic corrosion rate for a given atmospheric environmental condition. The model predictions are compared with experimental data in the literature. Good agreements are observed. The model is further used for prediction of galvanic corrosion of two identical vehicles located in two different geographic locations (i.e., Miami Beach in Florida and Wendover in Nevada) in the year 2021 leveraging weather station data. Additionally, a Bayesian estimation method is used to account for uncertainties in the model parameters and estimation of the probability of failure.