Numerical simulation of the macrostructure evolution of a heavy steel ingot

The macroscopic grain structure of a 234 t steel ingot was simulated by using cellular automaton and finite element method provided by ProCAST. The growing speed coefficient of Kurz-Giovanola-Trivedi model was modified to reflect the growing process of grains. Chilled fine grains, columnar grains and coarse equiaxed grains were obtained from surface to core of the ingot, which were roughly in agreement with the measured ones. The influences of the nucleation and growth parameters were investigated. It was found that the relationship between nucleation and growth determines the macrostructure profile. An index, the ratio of growing speed over the nucleated sites at certain undercooling was proposed to estimate the macrostructure tendency.