按照传统的铸造模式，大型铸锭中严重的成分偏析和晶粒尺寸差异是不可避免的。为了提高大型铸锭的品质，本项目提出一种基于成分差异微包浇注的大型铸锭浇注方法。该方法通过减少单次参与凝固的金属液的量，逐次使各次凝固的金属堆积在一起形成大型铸锭，通过调整单包浇注量、成分以及浇注间隔时间，控制金属液的流动状态、提高后凝固区域的冷却速度，从而降低宏观偏析、提高晶粒尺寸与形态的均匀性。构建考虑枝晶的柱状晶-等轴晶三相模型，耦合凝固过程的微、宏观现象；通过数值模拟优化单包浇注量、成分和浇注时间间隔等实验参数，获得偏析形态和晶粒分布图像；将微包浇注模拟、实验结合并与大型铸锭生产数据进行相互优化，揭示大型铸锭微包浇注过程中目标铸锭的成分偏析控制与单包浇注量、成分以及浇注间隔时间的关系，实现基于目标铸锭重量和品质要求信息的微包浇注程序，为大型铸锭的高品质铸造提供理论和技术基础。

The serious macrosegregation and grain size inhomogeneity in large ingots are unavoidable by conventional cast method. In this project, a new cast method which is based on the different-concentration multi-ladle pouring technology is proposed in order to improve the quality of large ingots. In this method, the amount of liquid metal solidified at one time is decreased. The solidified liquid metal layers accumulate successively and compose a large ingot. The liquid metal flow state is controlled and the cooling rate of the later solidified region is increased by adjusting the amount of liquid metal, the composition and the time interval for each pouring. Consequently, the degree of macrosegregation will be decreased, and both the grain size and morphology will become more homogeneous. A dendritic columnar-equiaxed mixed model which couples both the micro and macro solidification phenomena will be established. The amount of liquid metal, the composition, the time interval for each pouring and other experimental parameters will be optimized by numerical simulation. The macrosegregation, grain size and grain morphology distributions can be predicted by numerical simulation. This project can 1) reveal the relationship between the macrosegregation of the target large ingots and the amount of liquid metal, the composition and the time interval for each pouring in the multi-ladle pouring processing; 2) achieve the multi-ladle pouring procedure which is based on the weight and quality requirement of the target ingot; 3) provide both theoretic and technical fundamentals for high quality cast of large ingot.