Rapid dendritic growth investigated with artificial neural network method

Rapid dendritic growth of gamma-(Ni, Fe) phase, beta-CoSb intermetallic com pound and alpha-Fe phase was realized by undercooling Ni-10%Fe single phase alloy, Co-60.5%Sb intermetallic alloy and Fe-40%Sn hypomonotectic alloy to a substantial extent. Their experimentally measured dendrite growth veloci ties were 79.5m/s, 12m/s and 0.705m/s, corresponding to undercooling levels of 303K(0.18T(L)), 168K(0.11 T-L) and 219K(0.15 T-L) respectively. Since t he usual dendrite growth theory deviates significantly from reality at grea t undercoolings, an artificial neural network incorporated with stochastic fuzzy control was developed to explore rapid dendrite growth kinetics. It l eads to the reasonable prediction that dendritic growth always exhibits a m aximum velocity at a certain undercooling, beyond which dendrite growth slo ws down as undercooling increases still further. In the case of Fe-Sn monot ectic alloys, alpha-Fe dendrite growth velocity was found to depend mainly on undercooling rather than alloy composition.