Fatigue Life Prediction and Reliability Assessment of CFRP Adhesively Bonded Joints in Offshore Wind Turbine Blade Applications: A Physics‐Informed Data‐Driven Approach

ABSTRACT Adhesively bonded joints made of carbon fiber reinforced polymer (CFRP) are critical structures of offshore wind turbine blades, yet they are particularly susceptible to fatigue failure influenced by marine environmental factors. Current methodologies for predicting the fatigue life and assessing the reliability of these bonded joints face limitations due to insufficient data, modeling complexities, and inefficiencies. This study introduces a novel, physics‐informed, data‐driven approach to analyze the fatigue performance of CFRP adhesively bonded joints subjected to multi‐environmental stress, thereby improving fatigue life predictions and reliability assessments. Initially, an innovative method that integrates a physics‐informed data‐driven framework for predicting adhesive fatigue life and modeling reliability is proposed, utilizing the cyclic cohesive zone model (CCZM) alongside single‐hidden layer feedforward neural networks (SLFN). Multi‐environmental aging tests were conducted on CFRP‐bonded joints of varying dimensions, leading to the development of a physics‐informed fatigue analysis method based on CCZM. Furthermore, a reliability assessment and sensitivity analysis were performed to evaluate the impact of uncertainty factors. This research provides significant insights for the structural design and safety analysis of bonded structures in offshore wind turbine blades.