Fig. 7: Results from a machine-learned model of γ′ solvus temperature, and predictions for unexplored Co-based superalloys with high γ′ solvus temperature. | npj Computational Materials

Fig. 7: Results from a machine-learned model of γ′ solvus temperature, and predictions for unexplored Co-based superalloys with high γ′ solvus temperature.

From: Automated pipeline for superalloy data by text mining

Fig. 7

a γ′ solvus temperatures distribution for certain superalloys. b Root mean square error for model selection by 5-fold cross validation. c The predicted values compared to values extracted from our pipeline for the SVR.rbf model showing the behavior of the training and test datasets. d The measured and predicted γ′ solvus temperatures of 15 superalloys recently reported in 2020 and 2021 that were not part of our database. e The microstructure and DSC heating curve for alloy Co-36Ni-12Al-2Ti-1W-4Ta-4Cr. f The microstructure and DSC heating curve for alloy Co-36Ni-12Al-2Ti-1W-4Ta-6Cr. g The microstructure and DSC heating curve for alloy Co-12Al-4.5Ta-35Ni-2Ti.

Back to article page