LAUSR

Accurate creep life prediction is necessary for the evaluation of the remaining creep lives of in-service turbine blades and for the design of new turbine blades in aircraft engines. In this study, an integrated computational material engineering methodology for predicting the remaining creep life of in-service turbine blades was developed by taking a microstructural criterion and creep strain criterion into consideration, and combining artificial neural networks with a modified θ projection model to assess the service temperature, stress, degradation time, and existing creep strain. To explore the application of the method and verify its accuracy, the microstructural degradations at different locations of two directionally solidified superalloy DZ125 turbine blades, which were in-service for 300 h and 980 h in different engines, were characterized and quantified. Using these results, the remaining creep life of the microstructures at different locations of the blade was predicted. Finally, these creep life prediction results were experimentally verified using miniature creep test specimens. The development of this new method provides a reference for the design and service evaluation of turbine blades made of directional solidified and single-crystal Ni-based superalloys.