Patterns in Directionally Solidified Alloys

Understanding segregation pattern formation within the solidification structure of directionally solidified alloys is of vital importance for optimising their performance. Segregation patterns are not formed by coincidence, but develop from the constraints imposed by the processing environment and the material chemistry. The most important structure that develops during directional solidification is dendritic. The scale and regularity of a dendritic pattern determines the micro-segregation profile, the propensity for defect formation, and the mechanical properties of the material; thus, is important to study practically.

This work contributes to the current literature through the creation of an automatic feature recognition tool, DenMap, which is developed to identify dendritic core position within transverse cross-sections of directionally solidified alloys. DenMap is subsequently advanced with a novel supervised machine learning algorithm, Shape-Limited Primary Spacing (SLPS), that facilitates rapid, accurate, and efficient quantification of cellular and dendritic packing patterns in directionally solidified alloys, formed under different casting conditions. The results reveal the tendency to form hexagonally packed segregation structures under steady state conditions, while all other packing arrangements constitute a metastable state.

Using the SLPS algorithm, it has been demonstrated that packing pattern formation and local primary spacing can be employed to deduce dendrite tip growth kinetics. Further, the role of curved isotherms on inducing lateral macro-segregation parallel to a growing solidification front has been revealed. Isotherm curvature results in non-uniform liquid compositional gradients developing parallel to a growing solid interface, leading to the formation of metastable packing, low-angle grain boundaries, and porosity formation within the microstructure. Using crystallographic data from time-of-flight energy-resolved neutron imaging and novel Bragg-dip post processing, it is established that lateral macro-segregation induces mosaicity within single crystals, where fastest growing dendrites demonstrate greater deviation of <001> from the growth axis.