Modeling of inclusion removal in a tundish

Mathematical models have been developed to predict the removal of alumina i nclusions from molten steel in a continuous casting tundish, including the effects of turbulent collisions, reoxidation, flotation, and removal on the inclusion size distribution. The trajectories of inclusion particles are t racked through the three-dimensional (3-D) flow distribution, which was cal culated with the K-E turbulence model and includes thermal buoyancy forces based on the coupled temperature distribution. The predicted distributions are most consistent with measurements if reoxidation is assumed to increase the number of small inclusions, collision agglomeration is accounted for, and inclusion removal rates are based on particle trajectories tracked thro ugh a nonisothermal 3-D flow pattern, including Stokes flotation based on a cluster density of 5000 kg/m(3) and random motion due to turbulence. Steel samples should be taken from as deep as possible in the tundish near the o utlet and at several residence times after the ladle is opened, in order to best measure the Al2O3 concentration entering the submerged entry nozzle t o the mold. Inclusion removal rates vary greatly with size and with the pre sence of a protective slag cover to prevent reoxidation. The random motion of inclusions due to turbulence improves the relatively slow flotation of s mall inclusions to the top surface flux layer. However, it also promotes co llisions, which slow down the relatively fast net removal rates of large in clusions. For the conditions modeled, the flow pattern reaches steady state soon after a new ladle opens, but the temperature and inclusion distributi ons continue to evolve even after 1.3 residence times. The removal of inclu sions does not appear to depend on the tundish aspect ratio for the conditi ons and assumptions modeled. It is hoped that this work will inspire future measurements and the development of more comprehensive models of inclusion removal. These validated models should serve as powerful quantitative tool s to predict and optimize inclusion removal during molten steel processing, leading to higher quality steel.