Uncertainty quantification of mechanical behavior of corroded Al-Fe self-pierce riveting joints with statistical shape modeling

Due to its versatility, efficiency, and rapid cycle time, the automotive industry widely employs Self-Piercing Riveting (SPR) as a preferred technique to join dissimilar materials. Despite its benefits, these joints may experience galvanic corrosion due to the interaction of two distinct materials in the presence of an electrolyte. The varying equilibrium potential of the metals triggers the occurrence of corrosion, potentially undermining the durability and safety of the joints. In the presented study, we have focused on estimating uncertainty on the Al-Fe SPR joint's mechanical behaviors due to the impact of the different material losses of galvanic corrosion. We start by conducting finite element (FE) simulations on the Al-Fe SPR joints for the material loss and shape change due to galvanic corrosion. Considering the limitation of the small quantity of the Al-Fe SPR joint FE model results due to computational/time costs, we propose to use the statistical shape modeling method to regenerate new images based on the existing image data. Secondly, these images are imported into a Multiphysics-based corrosion simulation platform to reconstruct the shape of the Al-Fe SPR joint and the mechanical behavior of the joint with different material loss of corrosions can be obtained through finite element analysis. Thirdly, an uncertainty quantification study is conducted to understand the statistical characteristic of the mechanical behavior of the Al-Fe SPR joint from new and existing simulation images. The case study results show that the statistical characteristic of mechanical behavior from both source data are similar, and the statistical shape modeling method could be helpful in situations with insufficient sample data for uncertainty quantification.