Fatigue life estimation of open-hole cold-extrusion strengthened structures using continuum damage mechanics and optimized machine learning models

In the aerospace industry, many structural components in aircraft use open-hole structures, which are highly susceptible to fatigue failure, thus reducing the service life of the aircraft. Relevant studies both domestically and internationally have found that the fatigue life of open-hole structures in aircraft can be enhanced by employing the cold extrusion strengthening process. This paper investigates the impact of the hole cold-extrusion strengthening process on the fatigue life of open-hole structures. Using the framework of Continuum Damage Mechanics (CDM), a life prediction model is developed to estimate fatigue crack initiation. Model parameters are calibrated using experimental data. Numerical simulations are conducted to study the residual stress distribution resulting from varying levels of interference, and the trends are analyzed. The structure's fatigue life is then predicted to identify the optimal interference level and understand the underlying mechanism of the cold-extrusion process. Additionally, a CDM-based machine learning model is developed, incorporating K-Nearest Neighbor (KNN), Gradient Boosting Regression Tree (GBRT), and Artificial Neural Network (ANN). Through comprehensive analysis, the optimal parameters for each algorithm are determined, enabling accurate fatigue life prediction while significantly reducing computation time.