Predicting creep failure life in adhesive-bonded single-lap joints using machine learning
Authors:
Open Access
- 26 February 2025
- journal article
- research article
- Published by Springer Nature in Scientific Reports
- Vol. 15 (1) , 1-17
Abstract
Accurately predicting the creep failure life of adhesive joints, particularly single-lap adhesive joints (SLAJs), remains still a significant challenge, requiring substantial time and resources and the ability to predict the duration of creep failure in SLAJs is critical to ensuring structural integrity and reducing the failure of creep-prone adhesive joints. In this study, machine learning (ML) was used to identify the critical features that ultimately influence the durability of SLAJs due to creep. These key features were determined through correlation analysis and sequential feature selection. Multiple ML algorithms were employed to analyze complex relationships among key features and predict creep failure life. Finally, the results of the analysis highlight the importance of features such as SLAJ creep strain, adhesive tensile strength (UTS), SLAJ creep stress, adhesive surface area (A), and Young’s modulus (E). Of the ML models tested, the random forest (RF) model was the most effective in predicting creep failure life. Moreover, the accuracy of the predictions made by the proposed ML model, using original code written in Python, has been verified in experimental tests. All datasets generated and analyzed during the current study, along with the code, are available in the repository accompanying the paper.Keywords
All Related Versions
Funding Information
- Grant of the Ministry of Science and Higher Education in Poland (0612/SBAD/362)
This publication has 45 references indexed in Scilit:
- Influence of Temperature, Humidity and Load Coupling on Mechanical Properties of Adhesive Joints and Establishment of Creep ModelPolymers, 2023
- Effect of creep on the mode II residual fracture energy of adhesivesJournal of Applied Polymer Science, 2021
- Experimental and numerical models to study the creep behavior of the unidirectional Alfa fiber composite strength by the photoelasticity methodMechanics of Time-Dependent Materials, 2021
- Predictions and mechanism analyses of the fatigue strength of steel based on machine learningJournal of Materials Science, 2020
- Gallium–Boron–Phosphide ($\hbox {GaBP}_{2}$): a new III–V semiconductor for photovoltaicsJournal of Materials Science, 2020
- A predictive machine learning approach for microstructure optimization and materials designScientific Reports, 2015
- Handbook of Adhesion Technology2011
- Cross-Validation MethodsJournal of Mathematical Psychology, 2000
- Mathematical representation of creep for high-temperature performance of nylon6.6 tire materialsJournal of Applied Polymer Science, 1999
- The relationship between minimum creep rate and rupture time in Cr-Mo steelsJournal of Materials Engineering and Performance, 1992