Fatigue strength reliability assessment of turbofan blades subjected to intake disturbances based on the improved kriging model

This paper aims to develop an efficient and precise reliability analysis method to enhance the numerical prediction accuracy for complex structures. Kriging, an implicit surrogate model, used to address highly nonlinear and complex problems. In this study, genetic algorithms (GA) are utilized to optimize the parameters of the Kriging model, which is then integrated with a distributed collaborative strategy to introduce the Genetic Algorithm Optimized Distributed Collaborative Kriging Model (DCGAK). Using the CFM56-fan blade as a case study, the impact of intake disturbances at the engine inlet is evaluated to assess the fatigue strength reliability of the blade. Comparison with different mathematical models demonstrates that the prediction accuracy of DCGAK closely aligns with the Monte Carlo sampling results, suggesting promising prospects for its application in numerical prediction and reliability analysis. This approach enriches the current methods for structural reliability analysis of complex mechanical systems.