A volume-averaging dendritic solidification model

Motivations The motivation of this study is to further develop and validate a multiphase solidification model that accounts for mixed columnar-equiaxed solidification with dendritic morphology, melt flow, and grain sedimentation. The model will incorporate the dendritic growth kinetics, which occur at the microscopic interfacial scale, with the global mass, species, momentum and energy transport phenomena, which generally occur at the process scale of engineering interests. This kind of solidification model is still not available, but is highly desired by the scientific community and the foundry industry. Methods For the first time we propose a five-phase dendritic solidification model based on a volume-averaging approach. The 5 phases (more precisely phase regions) are: the equiaxed dendrites, the interdendritic melt in the equiaxed grains, the columnar dendrites, the interdendritic melt enclosed between the secondary or higher-order columnar dendrites, and the extradendritic melt. The fundamentals of the volume averaging approach were described by Beckermann and co-workers in the 1990s and a treatment of the dendritic morphology was suggested by Rappaz and co-workers in the late 1980s. These works have been extended further by many researchers, including the authors` current contributions. In the current development a commercial CFD code, FLUENT, is used for solving the global transport phenomena. The microscopic interfacial kinetics are volume-averaged, and implemented into the FLUENT code via a user-programming interface. Model assumptions will be verified with available theoretical models as well as comparison with real castings (Al-Cu ingots). The as-cast microstructure and concentration distribution in the ingots will be measured and compared with the numerical predictions. Scientific impact and expected results " Knowledge about the influence of the grain sedimentation and melt convection on the as-cast macrostructure including CET (columnar-to-equiaxed transition) will be obtained. " Understanding of the formation of macrosegregation will be improved. " Better understanding of the formation of the interdendritic and extradendritic phases (e.g. eutectics) during solidification, an important scientific and engineering interest, will be gained. " Direct application of the (volume-averaged) model to industrial processes can be carried out to quantitatively predict the as-cast structure, grain size distribution, local interdendritic and extradendritic (eutectic) phase distribution, macrosegregation, etc. The intermediate and final research results will be published in peer-reviewed journals, and presented at international conferences.

At some stage in the production of every metal part the metal material has to be melted and solidified to form the as-cast structure. The service performance of the metal part, i.e. its mechanical properties, is strongly dependent on the as-cast structure. Although casting of the metal material is an old art/technique with thousand years of history, many fundamental issues behind the formation of the as-cast structure still remains mystery and to be understood. For example, how does the growing columnar dendritic structure (in the form of fir tree) interact with the melt flow? How do the moving/settling equiaxed crystals (in the form of the snowflake) interact with and influence the growth of the columnar dendrites? What is the final as-cast structure: pure columnar, pure equiaxed, or a mixed columnar-equiaxed structure? A numerical model to answer above questions is highly desired by both scientists and foundrymen for understanding the physics behind solidification and for an optimal control of the production process. The motivation of this study is to refine and validate a five-phase mixed columnar-equiaxed solidification model, as previously proposed by the project proponent, with the aim to predict the as-cast structure. The model has incorporated the dendritic crystal growth kinetics with the global mass, species, momentum and energy transport phenomena. Firstly, the model was verified to be able to reproduce the solidification process and the structure formation of different laboratory experiments: (i) the unidirectional solidification of the transparent alloy Neopentylglycol-(d)Camphor under well-controlled microgravity (no flow) condition; (ii) the flow-solidification interaction in the transparent solution of NH4Cl-H2O; (iii) the casting trials of Al-Cu alloy ingots. Secondly, the modeling results help to explore most necessary details of solidification physics behind the experimentally verified facts, and new interesting knowledge was obtained in following aspects: (i) transport of the equiaxed crystals and its impact on the as-cast structure formation; (ii) macrosegregation mechanisms by natural convention and crystal sedimentation; (iii) columnar-to-equiaxed transition (CET) under diffusive crystal growth condition; (iv) sensitivity of the as-cast structure to the crystal dendritic morphology. Most recent findings were published in world-renowned scientific journals and presented in international conferences.