Research on low cycle fatigue reliability-based robust design optimization of turbine blade

To solve the problem of low cycle fatigue reliability-based robust design optimization of turbine blade, a series of fatigue tests of the blade material under high temperature were completed. According to the tests data, the randomized approach of deterministic equation was introduced to obtain the probabilistic strain-life curves of the blade material. The Bezier curve was employed to establish the blade model line and the parameter models of the turbine blade and flow field. After, the thermal-aero-structure coupling finite element method was used to obtain the aerodynamic efficiency, stress and strain distribution of the blade. Then, the fatigue reliability-based robust design optimization model was established and the response surface method was used to simulate the output performance functions and the limit state functions of the blade. Finally, the design optimization result was calculated with the sequential quadratic programming method by taking the fatigue reliability as the basic constrain condition. The results prove that: the low cycle fatigue reliability and robustness of the blade are improved remarkably after design optimization; the model and the methods are correct and can be served as references for the low cycle fatigue reliability-based robust design optimization of turbine blade and other complex structures. © 2013 Chin. Soc. for Elec. Eng.