Search Results

The process parameters to manufacture structural aluminum alloys are critical to their ductility. In particular, quench rate after solution heat treatment impacts the extent of grain boundary precipitation and the formation of precipitate free zone (PFZ) during later artificial aging. Cu-containing 6XXX alloys used for high strength automotive applications are quench sensitive as the Cu addition leads to Q-phase precipitation at grain boundaries, resulting in loss of ductility, which can negatively affect downstream manufacturing steps such as automotive joining and forming processes. Self-piercing rivet (SPR) joining, is a single step, spot joining process used to mechanically connect sheet materials together in automotive body structures. Ductility has been identified as an important metric of material rivet-ability or the ability to make a successful, crack-free SPR joint.

Self-piercing rivet (SPR) joints are a key joining technology for lightweight materials, and they have been widely used in automobile manufacturing. Manual visual crack inspection of SPR joints could be time-consuming and relies on high-level training for engineers to distinguish features subjectively. This paper presents a novel machine learning-based crack detection method for SPR joint button images. Firstly, sub-images are cropped from the button images and preprocessed into three categories (i.e., cracks, edges and smooth regions) as training samples. Then, the Artificial Neural Network (ANN) is chosen as the classification algorithm for sub-images. In the training of ANN, three pattern descriptors are proposed as feature extractors of sub-images, and compared with validation samples. Lastly, a search algorithm is developed to extend the application of the learned model from sub-images into the original button images.