A rapid multidisciplinary life optimization method for turbine blades with a large number of film cooling holes

To prevent turbine blades from overheating, a large number of film cooling holes are engineered on the blade body. Under the thermal/centrifugal/aerodynamic loads, film cooling holes are prone to failure due to the stress concentration near the holes. Owing to the high computational cost and the significant number of optimization variables, it is challenging to achieve the multidisciplinary optimization of film cooling holes, with the objective of extending life. In this paper, a reduced-order model is proposed to substantially improve the computational efficiency of physical fields distributed on turbine blades. Five types of modules are developed, including convective cooling/film cooling/heat conduction/planar stress/life prediction. The analytical solutions and empirical relationships are incorporated into each module to reduce the computational costs. To accelerate the convergence of optimization, the structural characteristic matrices are constructed to substantially reduce the number of optimization variables, and the typical geometric features of a large number of holes can still be retained. Combined with the high-throughput simulations facilitated by reduced-order model, the genetic algorithm is utilized to optimize the film cooling holes. The proposed method is applied to a typical turbine blade with a large number of holes. After optimization, the creep/fatigue lifetime is increased by 2.33 times. Finally, the conventional simulation method is utilized to validate the results. The reduced-order model can reduce the computational time by 3 orders of magnitude, compared with conventional method. The relative errors are within 3.5 % and 10.8 % for the simulation of temperature and stress at feature points. Based on the simulation of optimized configuration using conventional method, the effectiveness of life optimization is validated.