Analysis of gas turbine blade low-cycle fatigue life and reliability
To ensure the safe and reliable operation of gas turbines,an analysis was conducted on a specific turbine blade to obtain the stress and strain distribution under rated conditions.A high-precision surrogate model was built using a deep fully-connected neural network to assess the impact of multiple sources of uncertainty on the low-cycle fatigue life and reliability of the turbine blades.The results indicate that the maximum stress in the turbine blade is 1 024.91 MPa,located in the middle of the suction side of the blade root platform.The mean life of the blade under material and manufacturing stochastic factors is 1.237 × 104 cycles,while under operational stochastic factors,it is 1.146 × 104 cycles.The design life of the blade was set at 8.0 × 103 cycles,with a reliability of 0.945 2 under material and manufacturing influences,and 0.936 8 under operational influences.The findings demonstrate the effectiveness of the proposed method in the reliability analysis of turbine blades,providing reference data for the design optimization,performance enhancement,and life management of gas turbine turbine blades.
gas turbinebladelow-cycle fatigue lifereliabilitydeep-learning neural network(DNN)
万方数据
维普
