Abstract
We aim to provide a study of material selection at the upper adherend subject for optimization of carbon fiber reinforced polymer (CFRP) adhesive-bonded joints with details on the stress distribution in the complex tri-material joint structure. The structure consists of a changing upper adhesive and CFRP lower adhesive, all bonded by a nano-thickness resin adhesive layer. The research, which analyzes almost 100 upper adherend substrates, hopes to answer how they influence stress distribution at the apex of a joint, a critical factor in bond strength. These results are essential in selecting the donor properties of the upper adherend in CFRP bonded joints. As this study also supports engineers and researchers in devising optimized machine learning models for addressing CFRP-bonded joint challenges, the accuracy of stress prediction is improved by applying machine learning techniques to the collected data more refinedly.