On the basis of a multi-length scale modeling, a mixture-averaged multi-component/multiphase micro- segregation model was proposed without pre-set function for the micro-scale solute profile. The model explains the effect of morphologies of solidifying phases and solid back diffusion （SBD） on segregation, and covers the two limiting solidification cases of Scheil and Lever-rule models. A commercial Thermo-Calc software package/database was linked to the algorithms via its TQ6-interface for instantaneous determination of the related thermodynamic data of the multi-component alloys. The influences of cooling rate and other parameters on the solidification path and micro-segregation behavior were numerically investigated by sample calculation of the ternary AI-Cu-Mg alloys. A parallel experimental investigation on AI-Cu-Si alloys solidified under different cooling conditions was conducted to validate the theoretical model. Reasonable agreements were gained between the predicted solidification paths and the measured results.
