Machine learning-based strength prediction of nano-reinforced adhesive and hybrid joints under hygrothermal conditions

Abstract

Listen

Translate article

This research focuses on the impact of adding fullerene and Single-Walled Carbon Nanotubes (SWCNT) on the strength of bonded and bonded/bolted joints. The study examines composite-to-composite (CTC) and composite-to-aluminum (CTA) substrates under three-point bending before and after hygrothermal aging. Samples were categorized into neat specimens, specimens with added fullerene, specimens with added SWCNT, and specimens with a combination of 50 % SWCNT and 50 % fullerene. Experimental results reveal that the optimal nanoparticle ratio for bonded joints differs from that for bonded/bolted joints. Nanoparticles in the adhesive matrix significantly reduced degradation during hygrothermal exposure, preventing interfacial debonding and slowing strength reduction. Mixed formulations notably improved cohesive strength, shifting failure from the adhesive interface to within the adhesive layer. This shift indicates enhanced joint performance and durability under both normal and aged conditions, with cohesive failure being the most advantageous for joint resilience. Six machine learning models (Ridge Regression, Decision Tree, Random Forest Regressor, Gradient Boosting Regressor, Support Vector Regression, Neural Networks) were utilized to predict the static strength of these joints, highlighting that joint type, substrate, nanoparticle type and percentage, and environmental aging significantly affect adhesive performance. The best models for bonded joints were Support Vector Regression and Decision Tree, while Ridge Regression and Gradient Boosting Regressor excelled in bonded/bolted joints. Neural Networks consistently underperformed. This research performed a thorough analysis of the correlation between machine learning-predicted samples and experimental samples, providing valuable insights into the aging and durability characteristics of bonded and bonded/bolted joints, particularly dissimilar joints (composite to metal). The findings contribute to the development of methods to optimize joint durability and minimize failure risks during operational use.