A two-stage predicting model for γ′ solvus temperature of L12-strengthened Co-base superalloys based on machine learning

As one of the candidate materials of the next generation aircraft engines, L1(2)-strengthened Co-base superalloys have drawn lots of attentions. However, Co-base superalloys have some disadvantages, such as gamma ' precipitates in the superalloys are metastable. Moreover, improving this superalloy through traditional experimental approaches is extremely costly and inefficient. Thus, it is necessary to develop a new approach that could make rapid and accurate predictions of the properties of the L1(2)-strengthened Co-base superalloys. In this study, the gamma ' solvus temperature, which is the basic property of L1(2)-strengthened Co-base superalloys, is predicted based on our two-stage approach. Firstly, the existence of the gamma ' precipitates are predicted. And then, the solvus temperatures of the candidates which are predicted to have gamma ' precipitates are calculated by our models. A new superalloy with high gamma ' precipitates solvus temperature is designed successfully with the help of our approach. The time cost of this approach is less than that of the traditional experimental approach. This approach could also be used to discover L1(2)-strengthened Co-base superalloys with other desired properties.