This thesis aims to investigate both the two-dimensional and three-dimensional (sub)structure of the butterfly martensite (α’BF) morphology. In order to do this, multiple different types of heat treatments were applied to a Fe-25Ni alloy to find the optimal morphology consisting of a low density of α’BF that is surrounded by austenite(γ). Two dimensional analyses were performed by applying a combination of optical microscopy, scanning electron microscopy and Electron Backscatter Diffraction (EBSD). The α’BF morphology was found to nucleate and grow as the first martensite (α’) morphology just below the martensite start temperature (Ms) within this 25Ni alloy. A decrease in austenisation time resulted in smaller γ grain sizes. This reduction in γ grain size resulted in a reduction of the Ms of the alloy which in its turn reduced the amount of undercooling applied to the material below Ms upon quenching towards room temperature, causing a decrease in the freshly formed α’ fraction. Upon observation of the freshly formed α’ it was found that the α’BF seemed to prefer formation near the centre of γ grains instead of near γ grain boundaries. When α’BF was formed near a γ grain boundaries, one wing tends to aligned itself with this boundary. Through trace analysis, it was found that the habit planes of α’BF were close to {557}γ, {225}γ and {3 10 15}γ, which are characteristic habit planes of lath, butterfly and lenticular α’, respectively. The orientation relationship (OR) between the γ and α’BF is found be a combination of both the Greninger-Troiano (G-T) and Nishiyama-Wasserman (N-W) OR. Three dimensional analysis was performed using both serial sectioning and 3D-EBSD. Serial sectioning showed that α’BF within this alloy was sensitive to formation upon mechanical polishing. 3D-EBSD gave insight on the three-dimensional morphology and substructure of the α’BF grains. It was observed that the junction plane could be non-continuous. Moreover, it is shown that the apparent wing angle greatly depends on the angle that the α’BF grain makes perpendicular to the sample surface and that the tail of the the α’BF grain can run along the entire length of the grain.

Bainite steels are in high demand in many application areas owing to their outstanding mechanical properties, mainly due to the presence of a combination of fine bainite plates and retained austenite. Understanding the complicated mechanism of bainite transformation is crucial to creating an optimum design. Researchers have addressed this challenge via computational modelling, where the transformation start temperatures of bainite (Bs) and martensite (Ms) are key indexes when designing highperformance steels and their heat treatments.

This thesis aims to explore the potential that data collection has to create a model, using experimental results of the bainite transformation process. A list of metallurgical accepted claims is elaborated to assess the quality of the data. Principal component analysis and clustering techniques are used to identify patterns, most important features and main relationships in the dataset. The results confirm the exponential carbon dependence that bainite and martensite transformation temperatures have, therefore requiring nonlinear models to predict them.

Following, regression models and machine learning algorithms based solely on the chemical composition are used to predict Bs and Ms. Train-test split series and cross-validation are used to evaluate the prediction and consistency of each model. The results show that the ensemble learning algorithms outperform the regression techniques. Random forest and gradient boosting decision tree provide excellent Ms prediction on the validation set with R2 values of 0.92 and 0.93. The smaller dataset size adds up to the complexity of bainite transformation, resulting in worse prediction models of Bs, where the random forest and gradient boosting decision tree R2 values are 0.68 and 0.67 respectively. Even though the models are showing signs of learning, the impact that outliers have demonstrates that the data is not good by itself to create a predictive model. The incorporation of microstructural and process parameters would provide significant advances to the models for designing bainitic steels.