Columnar dendritic growth during directional solidification of neopentylglycol-(d)camphor investigated under microgravity conditions

• ISS provides microgravity conditions for convection-free solidification experiments. • Columnar dendritic and eutectic growth in transparent system Neopentylglycol-(d)Camphor was investigated. • Dendritic undercooling and dendrite spacing depend on experimental conditions, composition and crystallographic orientation. • Supercooling of interdendritic eutectic phase followed a linear relationship with cooling-rate. We investigated dendritic solidification in the binary transparent organic model alloy neopentylglycol-(d)camphor under microgravity conditions onboard the International Space Station. Three cartridges containing alloys with hypoeutectic compositions of 0.2, 0.3 and 0.375 wt-frac. (d)Camphor were processed repeatedly in the “Transparent Alloys” solidification facility. In these Bridgman-type experiments, the dendritic grain structure − columnar, equiaxed or both mixed − was determined by in-situ optical observation as a function of alloy composition, crystallographic growth orientation, applied temperature field and cartridge withdrawal velocity. In most experiments, a columnar-equiaxed transition was achieved by increasing the cartridge velocity. We describe the experimental setup and procedure and focus on analysis of the columnar dendritic growth regime here. This includes estimates of the temperature gradient and supercooling at the columnar and eutectic front in combination with thermal modelling, as well as evaluation of a characteristic dendrite spacing. The results complement ground-based and other short-duration microgravity experiments in the same compositional range of the alloy. For eutectic growth, the undercooling increases linearly with cooling-rate. For columnar dendritic growth, the comparison of the kinetic law with previous experimental results exhibits similar trends but also significant deviations, which may be attributed to the different experimental set-up, data analysis and accuracy of temperature prediction. The determined characteristic dendrite spacing is well predicted by the models of Trivedi, Lu/Hunt and Ma/Sahm.