Macrosegregation is a critical defect in industrial steel ingots in both conventional ingot castings and continuous castings. Improvements in terms of macrosegregation require to be able to precisely quantify the intensity and the repartition of macrosegregation along the cast product. However, traditional macrosegregation measurements are a tedious and time consuming process. In this study, three different product lengths of the same caster strand were taken during the same heat. Each sample was cut along the longitudinal central plane of the cast product. One additional transverse section was also cut for one of the sample. Drilling with constant spacing was performed on 50% of the area of longitudinal sections and 25% of the transverse section. Carbon and sulfur contents were then measured on each drilled location. Plotting methodologies to produce chemical maps are carefully described regarding the different parameters such as color map choices and normalization methods. Data for all the characterized sections is compared with each other and the differences are commented with the help of the process parameters.
Dans une piece metallique, des pores residuels non desires sont toujours presents apres les operations de fonderie. Pour refermer ces pores de taille generalement negligeable par rapport a la taille de la piece, les industriels utilisent des procedes a chaud tels que le forgeage libre ou le laminage. Dans la litterature, plusieurs modeles a champ moyen predictifs en terme d'evolution du volume des pores traitent de cette problematique. Recemment un nouveau modele a champ moyen, appele Cicaporo4, a ete elabore et continue d'etre ameliore afin de modeliser la refermeture des pores. Ce modele se base sur une approche multi-echelles2 utilisant des simulations sur Volume Elementaire Representatif (VER). La plupart des modeles classiques etudient la refermeture en fonction de la triaxialite des contraintes et de la deformation plastique equivalente. Pour l'elaboration du modele Cicaporo, les effets de la morphologie des pores et de leur orientation dans la matrice ont ete investigues.
Although a significant amount of work has been already devoted to the prediction of the macrosegration in steel ingots, the majority of studies considered the solid phase as fixed. It has been shown that with such an assumption it is not possible to predict the macrosegregation in the centre of heavy steel ingots. The motion of the equiaxed grains is generally suspected to be the cause of this macrosegregation. We developed a multiphase and multi-scale model, which is able to describe the evolution of the equiaxed crystals and to take into account their motion. Fragments of dendrites are assumed to be at the origin of the equiaxed grains. The flow of the interdendritic liquid when the grains pile up and are packed is also described. The model was numerically implemented using a 2D finite-volume formulation. We applied the model to study the development of the macrosegregation and the macrostructure in a 65 ton steel ingot that was experimentally analyzed. The ingot was cast with particular conditions. A metallographic study was performed and the macrosegregation pattern was characterized. The macrosegregation pattern predicted by the model is presented and compared to experimental results. It is shown that in the case of such a heavy ingot, the motion of the equiaxed grains is indeed the main cause of the negative macrosegregation in the centre of the product
